Louise Kammacher

What we talk about when we talk about mother tongue Linguistic minority pupils and the issue of mother tongue education in Copenhagen ISBN: 978-87-91621-96-3 ISSN: 0901-9731 EAN: 9788791621963 Københavnerstudier i tosprogethed Københavns Universitet Humanistisk Fakultet 2016 (Udgivelsen er bind 72)

Køn , identitet og kommunikation på arbejdspladsen

Narges Ghandchi

Narges Ghandchi

What we talk about when we talk about mother tongue Linguistic minority pupils and the issue of mother tongue education in Copenhagen

Københavnerstudier i tosprogethed 72

Københavns Universitet Humanistisk Fakultet 2016

Bind 72

What we talk about when we talk about mother tongue

Linguistic minority pupils and the issue of mother tongue education in Copenhagen

Narges Ghandchi

Copenhagen Studies in Bilingualism vol. 72 University of Copenhagen Faculty of Humanities Copenhagen 2016

Copyright 2016: Narges Ghandchi WHAT WE TALK ABOUT WHEN WE TALK ABOUT MOTHER TONGUE. LINGUISTIC MINORITY PUPILS AND THE ISSUE OF MOTHER TONGUE EDUCATION IN COPENHAGEN Copenhagen Studies in Bilingualism vol. 72 ISBN: 978-87-91621-96-3 ISSN: 0901-9731 This study was supported by the Super-MOTE project, funded by the Danish Council for Independent Research. The Copenhagen Studies in Bilingualism is a series published by the Faculty of Humanities at the University of Copenhagen Editorial board: Juni Söderberg Arnfast, Centre for Studies in Danish as a Second and Foreign Language, University of Copenhagen Anne Holmen, Centre for Internationalisation and Parallel Language Use, University of Copenhagen Janus Spindler Møller, LANCHART, University of Copenhagen Jakob Steensig, Section of Linguistics, Aarhus University Distribution: University of Copenhagen, Faculty of Humanities This volume is edited by: Juni Söderberg Arnfast and Anne Holmen

2

TABLE OF CONTENTS LIST OF APPENDICES .............................................................................. 5 ACKNOWLEDGEMENTS .......................................................................... 6 1. INTRODUCTION .................................................................................. 7 1.1. Aim of the study ....................................................................... 10 1.2. Background of the study ......................................................... 11 1.3. Outline of the study ................................................................. 12 2. THEORETICAL FRAMEWORK ............................................................ 14 2.1. Introduction .............................................................................. 14 2.2. Ideological perceptions of language and language users .. 15 2.2.1. Language ideologies and the issue of language ......... 15 2.2.2. Native speaker, mother tongue language user, and heritage learner .......................................................................... 18 2.2.3. MT in the Danish context................................................ 22 2.3. Space, territory and language ................................................ 24 2.3.1. Territoriality and territorial restrictions........................ 25 2.3.2. Territories within interactional situations .................... 28 2.4. Language registers within geographical and interactional territories .......................................................................................... 29 3. SOCIO-CULTURAL AND LANGUAGE-IDEOLOGICAL BACKGROUNDS OF TURKISH, IRANIAN AND ARAB MIGRANTS ...... 33 3.1. Demography of the migrant groups with Turkish, Iranian, and Arabic backgrounds ................................................................. 34 3.2. Socio-cultural background of the three languages: Arabic, Farsi, and Turkish ............................................................................ 35 4. METHOD AND METHODOLOGY ...................................................... 39 4.1. Fieldwork study ........................................................................ 40 4.2. Fieldwork sites .......................................................................... 42 4.3. Participants and data ............................................................... 44 3

4.3.1. Participants ....................................................................... 44 4.3.2. Data, devices, and data process .................................... 48 5. ANALYSES ........................................................................................... 51 5.1. Analysis (I).................................................................................. 51 5.1.1. Whose mother tongue wherein? ................................... 51 5.1.2. What language in whose country? ................................ 73 5.2. Analysis (II) ................................................................................ 82 5.2.1. “Normal Farsi” .................................................................. 83 5.2.2. Turkish varieties in Turkish MT class ............................ 87 5.2.3. Standard Arabic ............................................................... 95 6. DISCUSSION AND CONCLUDING REMARKS ................................102 REFERENCES ........................................................................................109 RESUMÉ ................................................................................................120 APPENDIX (1): TRANSCRIPTION CONVENTIONS .............................122 APPENDIX (2): UNI-PERS-ROMANIZATION SYSTEM FOR FARSI .....123 NOTES ...................................................................................................124

4

LIST OF APPENDICES

Appendix (1)

Transcription Conventions

Appendix (2)

Romanization system for Farsi

5

ACKNOWLEDGEMENTS I would like to express my sincere gratitude to a number of people without whom this study would not have been possible. First, Martha Sif Karrebæk who supported my work at several levels: she contributed indefatigably to the main project upon which this study is built and she encouraged the minor project of the present publication. She read this book - based on my master’s dissertation - in its different phases, and inspired me through her discussions over the course of the entire writing process. Secondly, I would like to thank the LANCHART Center that supported this publication in its earlier version. In particular, I wish to express my gratitude to Frans Gregersen. Thirdly, I am grateful for the support of the language researchers affiliated with two projects at the Institute of Nordic Studies and Linguistics: Mother Tongue Education for the linguistic minority pupils in the super-diverse metropolis of Copenhagen (Super-MOTE) and Copenhagen Studies in Everyday Languaging. I have learned from the academic insights of both groups in various ways and I have used the data the Mother Tongue researchers collected, transcribed, and processed in different stages of my investigations. Also, I owe a huge debt of gratitude to the late sociolinguist Jens Normann Jørgensen. And finally, my special thanks to the editors of this volume, Juni Söderberg Arnfast and Anne Holmen. I am solely responsible for any faults in this book. Narges Ghandchi Amager, June 2016

6

1. INTRODUCTION Når jeg til sidst får sagt farvel til klassen, møder jeg Smillas far (oprindeligt engelsk-iransk) ved skolens indgang. Jeg spørger ham om han har fået projektets folder. Smilla (13 år) hvisker noget, jeg ikke kan høre, på engelsk og trækker ham i armen på vej mod bilen. Jeg fortæller ham dog insisterende om modersmålsprojektet. Han smiler skævt og siger, i dette tilfælde er det ikke deres modersmål. Dette er deres ’fadersmål’. Jeg siger ”nej. Faktisk er der et andet begreb på engelsk til sådan en sag. Og det er ’heritage language’. Faren synes det er et interessant begreb, og tilføjer, at hans mor i mange år har ledet en skole for iranerne i Storbritannien”. [When I finally say bye to the class, I meet Smilla’s father (originally English-Iranian) by the entrance of the school. I ask him, whether he has received the project folder. Smilla (13 y.) whispers something unclear in English and pulls his arm towards the car. However, I insist on telling the father about the mother tongue project. He smiles wryly and adds, in the case of his daughters it is not their mother tongue. It’s their ‘father tongue’. I say, “no, actually there is another term in English for such cases. And that is ‘heritage language’. The father finds it an interesting term, and adds that his mother has been running a school for Iranians in Great Britain for many years.] (Field note by NG; 19-04-2013) The excerpt above is taken from one of my ethnographic fieldwork diaries, relaying a short conversation on which language one’s mother tongue is. It illustrates doubt raised by a father on mother tongue as a term, following my enquiry about the reception of the project’s flyer. Having a dual national background, the father sends his children to a particular class for instruction in the Farsi language. The children (3) are raised in a mixed-marriage family with Scottish and Iranian-British background. Their mother does not speak or understand Farsi. In the Farsi language learning class, 7

the children learn to write and read in Farsi, but are hesitant users of the language during my interactions with them in Farsi about everyday topics. Instead, they look more comfortable while communicating in either Danish or English. According to Danish school terminology, this type of language education is institutionally named “mother tongue education,” as ordained by the Danish Ministry of Education. The ethnographic project I present to the father also has mother tongue education as its focus. In this context and by this term, one might presuppose pedagogical instruction in languages that are considered minority languages, specific to certain children with a different linguistic background than Danish. One of the objectives of this book is to argue that the terminology is a problem in the context of such educational endeavors, and in the understandings around them. In the excerpt above, although I refer to Farsi as the mother tongue, the father ironically questions the established term “mother tongue,” seeing that Farsi seems to be more of a “father tongue” - considering the kinship link - in the case of his children. I suggest another widely used term: heritage language (Valdés 1999; Fishman 2001). The highlighted term of “mother tongue” addresses the question of what somebody’s relation to a “mother tongue” signifies and what understandings exist around this term in the “current globalization era”, as Blommaert (2010: 14) puts it. This era is characterized by flows of global migration alongside increasingly new modes of communication technologies across the globe which affect “the sociolinguistic patterns of language in society (…) and super-diverse patterns of urban multilingualism” (ibid). How can individuals’ mother tongue be defined and understood in relation to their mobility and new and prior places? To what extent may such a mother tongue be used or required within the context of mobility and migration? The issue is noticeable in two ways: On the one hand, English functions as a “hyper-central language” that attracts diverse speakers (see De Swaan 1993; Ammon 2006). On the other hand, knowledge of more languages is an occupational requirement in the long-term (Duchêne et al. 2013). How meaningful is it, then, to stress a language as one’s mother tongue and to maintain it outside the country of origin? And how are these issues represented,

8

understood, and negotiated in social as well as institutional contexts? Recalling the family in question in the above vignette, the family members were from three heritage backgrounds. The parents had been working as skilled migrant workers in Denmark since 2008. The children of the family attended an international school that taught in British English as well as Danish, as the father reported in a qualitative interview. Likewise, the languages his children (from 8 to 13) used during their school time were either English or Danish. Neither of the parents could speak Danish. The father only spoke Farsi occasionally to his children. The children communicated in a combination of Farsi-English with their paternal grandmother (originally Iranian, living in London). During my fieldwork inside the Farsi language class, I noticed the children shifted from Farsi to English and Danish when addressing different participants. Questions regarding such cases emerged, concerning the migration and mobility of people with different heritage backgrounds, which are the focal points of this book. I list some of the emerging reflections here: (1) What are the requirements for considering a language the mother tongue of a person? (2) What language is perceived as a person’s “mother tongue” and who are “mother tongue speakers”? (3) What are the understandings and beliefs of individuals who participate in mother tongue language classes? (4) What are the understandings concerning the relationships between one’s mother tongue and his or her country of origin? (5) How does language connect with place (e.g., country of origin, country of migration, country to which one’s parents once migrated) and with space (i.e., social and interactional spaces, which influence whether one speaks Danish, English, Arabic, and so forth)? I use the Danish institutionalized label of “mother tongue education” (henceforth, MTE) and the more generally available term MT when referring to the language learning instruction

9

specified for children with other languages than Danish (I return to this latter term below). Within this educational environment, such children learn the language of their or their parents’ countries of origin. MT courses also link certain social actors (i.e., participants) together: the MT school children, their parents (one or both of them), and their families, and the MT language teacher. Generally, the target language (MT) also links these people to a specific location; namely, an MT classroom. At the same time, as we see below, participants in this study ascribe an MT to particular and restricted places, such as “home” or “ordinary (Danish) school,” and to places outside of Denmark, e.g., Turkey. Participants treat these linkages as unexceptional; it seems to be a naturalized assumption that having Turkish as one’s MT is an immediate characteristic of being originally from a country where Turkish is the official language. 1.1. Aim of the study Methodologically, the study is part of a larger linguistic ethnographic project (Maybin & Tusting 2011; Rampton et al. 2015) in and around four MT classes in Copenhagen. In this study, I concentrate on one Turkish, two Arabic, and two Farsi classrooms. This study benefits from the methodological and analytical tools related to ethnographic studies. Departing from ethnographic studies in education (Blackledge & Creese 2010; Weber & Horner 2012), I explore the issue of MT through a combination of three strands: The first focuses on various understandings of MT. How do different social actors of the study approach the issue of MT? How do they express their correspondence to an MT language as a natural, intrinsic, and beyond discussion relation? (see 2.2.2.) The second strand focuses on places and social/interactional spaces within which MT is considered preferred or dispreferred among the participants (see 2.3.). Apart from varied linkages between MT and the spaces to which social actors refer, this study is also concerned with linguistic ideologies and attitudes (Silverstein 1979; Woolard 1992; Schieffelin et al. 1998). Herein lies the third strand of the study: What are the ideologies represented in and around the MT classes? Through 10

linguistic ideologies, I refer to sets of beliefs concerning language and language use. This includes to-do and not-to-do lists of language use (Blommaert et al. 2005; Blommaert 2007) and values. The participants in the study used standard language or other varieties of MTs. My major aim with this focus is to shed light upon norms and ideologies and to investigate how they influence and co-construct a particular space known as MT classes. In summary, I explore the participants’ understandings of MT, spaces, and ideologies to answer the following question: Which space do language users establish through the use of mother tongue? 1.2. Background of the study My study is situated within a particular social, institutional, and institutionalized context, namely MT classes. MTE is a particular type of institution in Denmark in which certain MT classes are organized and offered in the extension of the school’s ordinary curriculum and schedule. This study is part of a larger ethnographic project with four MT foci: Arabic, Farsi, Turkish, and Polish. The project motivated me to investigate MT classes. I was a participant observer (Spradley 1980) in the Farsi classrooms (201314), and I also attended the Turkish classes as a passive observer on select occasions (Spradley 1980). In the Farsi classes, I - being a 1st generation Iranian migrant - partiallyi participated in the instruction as a volunteer assistant teacher. In the Turkish site, I had limited ability to establish connection with the field. In the case of the Arabic classes, I did not observe the classes as three other researchers already participated in the sessions. For this study, I had access to the other researchers’ data. The three teachers shared some similar experiences with and understandings of the target languages, of the children’s patterns of language use, and of the spaces and places to which the target languages were related. It was my initial idea to reflect on these observations. I have a second interest in studying these three classes, which deals with the intertwined historical and socio-cultural background of the countries of origin of the three MTs. Iran and Turkey share simultaneous projects of modernization (early 1920s) 11

and experienced top-down attempts to establish language policies within nation-state settings (Keyman & Yilmaz 2006). The Arabic countries similarly have a tradition of socio-political modernization, language regimes, and complex attitudes toward Western countries, foreign words, and nationalism. This is further detailed below. 1.3. Outline of the study This book is divided into two main parts, a theoretical and an analytical, and consists of 5 chapters. I present my theoretical framework in chapter 2. This chapter contextualizes the book within three theories: (1) The definition of native language and mother tongue; (2) the socio-psychological theory of the territories of the self (Goffman 2010), along with social and geographical insights on territoriality (Sack 1986; Storey 2001); and (3) the sociolinguistic approach to language registers (Agha 2006; 2007). In this chapter, I define and briefly discuss definitions regarding ideologies of language use, language users, and spaces. I intend to examine how the theories could be applied in relation to the empirical data of this study. I also devote a shorter part (chapter 3) to the sociopolitical background of the languages in the study, with a focus on aspects of language politics in their related countries of origin. This stems from my belief that sociopolitical background affects the way languages are treated and positioned. The empirical data are introduced in chapter 4, where I present the ethnographic fieldwork, participants, and data collection and process too. Following chapter 4, I present my data analysis. This analysis falls in two parts, selected through the lenses of the study’s theoretical framework. The first part is related to the first two theoretical strands, with a focus on understandings and usage of MT. I also demonstrate how space adds to the meaning of an MT. The second part of my analysis concerns linguistic ideologies on the target languages taught in MT classes. In this section, I analyze the 12

data to examine how language adds to the meaning of space. Finally, this study will be concluded with a discussion of the research analysis and findings.

13

2. THEORETICAL FRAMEWORK 2.1. Introduction In order to analyze the MT institution in the present study, it is necessary to discuss the terminology used with reference to or relevant for the MT institution and MT speakers. Studying the usage of these terms institutionally, academically, and among individuals delineates the different understandings of them. In this chapter, I briefly present theoretical approaches that discuss the relationship between the terms as they are relevant to language – specifically, to language users and academic contexts (part 2.2.), as well as language users and space/place (part 2.3.). For this purpose, I discuss the terms language, native speaker, MT speaker, and heritage language (learner) (section 2.2.2.). Studies show that these terms carry an ideological load. Language ideology refers to an overall set of normative beliefs and attitudes with regard to language (Blommaert & Verschueren 1998; Ag & Jørgensen 2013; Weber & Horner 2012). I start with a description of language ideologies (2.2.1). This will be important in the remaining sections of the theoretical part as well as in the study’s data analysis. In a later section (2.3.3) I provide a review of a dominating, institutional understanding of MTE and related types of educational endeavors and initiatives in Denmark. The third part of this chapter (2.3.) treats two theories defining space (social and interactional space) and place (restricted geographical space (Mæhlum 2010)) in relation to individuals. Both presentations of theoretical terms aim to study how the institutionalized terms related to MTE are used by individuals in everyday language. The fourth part (2.4.) concerns a theory of another type of relationship between space, language, and language varieties, which can also be deemed as language ideological. I focus on what the three concepts of space, language, and language ideologies correspond to and how language users treat varied types of language use as related to specific places, and evaluate them in relation to specific spaces/places. I illustrate how certain language ideologies that distinguish between standard and other forms of a language (i.e., the variety termed “mother tongue” 14

in MT classes) influence what we understand, value or devalue, and label other types of language forms and other language types’ users. In order to do this, I describe the theory of language registers (Agha 2006; 2007), which will be the backbone of the second part of my analysis. This theory deals with cultural understandings and the labeling of speech forms associated with social practices and “[indexing] a stereotypic image of social personhood or interpersonal relationship” (Agha 2007: 80). 2.2. Ideological perceptions of language and language users 2.2.1. Language ideologies and the issue of language Language ideologies present language users’ beliefs about and attitudes towards a language (e.g., Danish) and its usage - either others’ or their own usage. Language ideologies are based on socio-cultural linguistic norms (Kroskrity 2000; Ag & Jørgensen 2013), which are co-constructed socially. Such norms influence language users to position one another in relation to certain language usage, particular places, and nationalities, with reference to understandings of competence, authority, correctness, and so forth (Agha 2007). Agha (2007: 126) divides norms (including norms of language) analytically into three levels: (1) they introduce externally observable correlations with a group of individuals and imply statistically recognizable and frequent patterns of behavior; (2) they involve reflexive models that normalize certain specific patterns of behavior; thus individuals come to consider such behaviors as normal; (3) the normalized patterns of behavior are codified as standards of appropriateness or correctness. In general, a language ideology is a “rationalization or justification of perceived language structure and use” (Silverstein 1979: 193). In addition, like other ideologies, language ideologies are supported by “commonsensical” ideas (Blommaert & Verschueren 1998) when speakers refer to them and are rarely subject to questions inside a social group (Blommaert & Verschueren 1998: 25). Prevalent language ideologies concern, for instance, language as a 15

separate entity, and thereby relate one national/official language to one nation and one national identity; or they distinguish between standard form and local varieties of a language and valorize them; they often highlight that speakers necessarily have a mother tongue (Skutnabb-Kangas & Phillipson 1989; Makoni & Pennycook 2007; Ag & Jørgensen 2013; Weber & Horner 2012). To illustrate this, I present a field note from the Farsi language class: A pupil (11 y.) named Shahin is introduced to the new word sarf (which means “grammatical inflection”), and I try to explain the word through German, which I know Shahin has some knowledge of. I do this because German is an inflection-rich language. I ask him what his first language is. “Farsi,” he answers. I respond: “Is Farsi your first language?” He says yes, and explains that he is Iranian. I do not go in detail and only continue and ask what his second language is, then. “Danish,” he answers. He later claims that German is his fourth language (Field note by NG; 01-02-2014). Shahin has positioned Farsi as his first language by virtue of the relation he makes between the language and Iran, his perceived country of origin or being Iranian. In establishing this connection, he rationalizes his choice of Farsi as his first language. This probably builds upon his socio-cultural experience (Kroskrity 2007) and language socialization. He is likely to have met this naming or the establishment of this linkage in his family or among other Iranian acquaintances. At the same time, his answer surprises me, as I have noticed during the classes that his expertise in Danish is higher than in Farsi, both in vocabulary and pragmatic usage. Regardless of this, he seems to perceive Farsi as his first language and Danish as his second. However, another possibility is that his perception stems from viewing Farsi as the target language of the “mother tongue class” and thus a language with a situationally higher status or relevance. Positions like Shahin’s reflect language ideologies; in this example, Shahin’s ideological beliefs situate his relation to different languages within a presupposed and perhaps normalized linguistic hierarchy. Individuals - like Shahin - may 16

naturalize connections between themselves as language users, their usage of language, and the country or the nation they are perceived or perceive themselves to have roots in. At the same time, the understanding that there are inherited and natural links between a language, a language user, and a nation builds on a historical tradition, and these ideas are grounded in a political project of defining social and national territories. Languages do not exist as real entities in the world and neither do they emerge from or represent real environments; they are, by contrast, the inventions of social, cultural and political movements (Makoni & Pennycook 2007: 2). In other words, the term language and the practice of naming languages are, historically and institutionally, inventions of a romantic era during which humanist traditions of nation-states (Appadurai 2000) postulated an intrinsic relation between one language (as an analytical entity), one nation, and one people. Among various ideological inventions, in the words of Makoni & Pennycook (2007), the terms “mother tongue” and “native speaker” have been conceptualized in relation to historical and social context. When UNESCO started to encourage MT instruction in 1953, the notion of MT was assumed to be “axiomatic[ally] the best medium for teaching a child (…) [through] which he learns more quickly (…) than through an unfamiliar linguistic medium” as noted in discussing the importance of language and education (UNESCO 1953: 11, my emphasis). MT was taken for granted as a language in which the child learned more quickly. In the later definitions stressing the importance and advantage of investing in and focusing on more than one language, MT was assumed to be the child’s first language, i.e., L1. L1 has been regarded as a language the child could speak at home before entering school (Ball 2010: 2). Shifts in the assumptions, definitions, and understandings have occurred over time following the global changes of socio-political and economic relations of language speakers and the places they have been and are related to. Perhaps the labels MT and MT 17

speakers are represented in the form of other/new denotations, and have received new associations in practice as a referential need. They may also be dependent on the realities of our time, as characterized by mass migration and mobility. In what follows, I will describe the two prevalent and ideologically based terms, native speaker and heritage language, from an academic perspective. Yet, when non-academic institutions and legal establishments use these academic terms and refer to them, this makes the understandings around the terms more complicated. Such institutional and legal references/entitlements regarding the usage and rights associated with the terms are taken up by mainstream society (e.g., municipalities, schools) and the migrated language users (e.g., ethnic associations) in order to teach or maintain the migrant languages across generations. Therefore, I will also look into how the term mother tongue is used from a Danish institutional point of view to discuss local (vs. global) understandings. 2.2.2. Native speaker, mother tongue language user, and heritage learner The concept of native language refers to one’s first-acquired language, and native speaker denotes a speaker of that language (Bloomfield 1933). The implicit assumption is that a native speaker has a range of intuitive knowledge of meanings, rules, general language skills, and creativity of language use (Stern 1983), or is perhaps even an “ultimate referee of linguistics” who can evaluate language performance (Jørgensen 2008: 154). Yet, Paikeday (1985: 87) states that the nativeness of a given language user does not involve the language of origin, and thereby suggests that a native speaker is a “proficient user of a specified language” irrespective of country of residence. The issue of native speaker has been much studied within the field of English language teaching (see below). As a theoretical concept, nativeness has been used as a baseline against which learners of diverse linguistic backgrounds are evaluated. On the other hand, schooling in MT in multilingual global (Blommaert 2010) environment seems to have added associations to the concept of nativeness when speaking of migrants and linguistic minority speakers. The right to use, maintain, and develop MT is regarded as a human right, and 18

policies are made and encouraged worldwide for the development of MTs among new generations of migrants and local minority languages (see for example Skutnabb-Kangas 1994; Ball 2010). Such policies aim partly to prevent first language erosion and attrition among migrants and minoritized language speakers. In the Danish context, for instance, MTE has been a type of environment where participants with different levels of language proficiency are positioned as MT speakers, and this has been the case since 1975 (Kristjánsdóttir & Timm 2007). However, there are different views to this within Europe, and speakers have been positioned as MT pupils, heritage language pupils (EACEA P9 Eurydice 2009), or pupils of complementary classes (Blackledge & Creese 2010). The common component in these labels is that such pupils are considered speakers of a migrant language as well as the language of their respective macro-society. Native speaker & MT language user Among a number of disciplines, the concept of native speaker has been studied in parallel to non-native speakers within Second Language Acquisition and specifically in Teaching English to Speakers of Other Languages (TESOL). The TESOL conceptualization historically classified students into two groups: native speakers and speakers with other linguistic backgrounds, i.e., “linguistic and social outsiders” (Leung et al. 1997). I find the newer TESOL approach with regard to nativeness interesting as it uses language proficiency rather than origin as a distinctive criterion. From the traditional TESOL perspective, a native speaker was an idealized term. A native speaker could think of and position himself as having certain rights regarding the language he had “ownership” of, when viewing language as a “package of features which comprise certain features and exclude all others” (Jørgensen et al. 2011: 31). In this sense, the term Second and Other Language learner was an equally idealized term with regard to the linguistic minority pupils of various types of language learning needs and linguistic backgrounds, all attempting to acquire English like a native speaker. In order to replace these rigid and idealized terms in relation to pedagogical organization and curriculum, Leung et al. (1997) propose that three criteria are crucial in examining the relationship between language and language users: language 19

expertise, language affiliation, and language inheritance. Language expertise is a way of looking at pupils’ language proficiency and thus their linguistic competence. Language affiliation concerns the ‘sense of affiliation’ language users have with the languages they have access to. This is a relevant concept both in cases where the language user is considered a native speaker and when he or she is not. Language inheritance is meant as an implicit critique of the automatic relation understood by the traditional approach of TESOL, in which language is a consequence of inheritance, birth, and family background. E.g., the MT-user Shahin in section 2.2.1 identified himself as a native speaker based on language affiliation and most probably also on inheritance relationship, whereas to Smilla’s father in the previous chapter, his children had a relationship to Farsi based on inheritance. From this perspective, it is possible to discuss the different forms of relation between language users and languages they use and choose. Therefore, moving from the rigid dichotomy of native/non-native speakers (with both linguistic majority and minority as group members), these criteria invite language educators and researchers to take such diversity into account with respect to the reality of actual language use. Heritage language user or learner Other perspectives on linguistic minority speakers have labeled them heritage language users. Although the term is not traditionally used in Denmark, I believe and show below it also carries some potential in the Danish context. Heritage language education was historically tied to the instruction of languages spoken by indigenous peoples and migrants, with reference to languages other than English in the USA (and English and French in Canada; see Cummins 1991; Duff 2008). I use Fishman’s (2001) definition as it is comprehensive and relevant (but see also Valdés 2001; Bale 2010). According to Fishman (2001), heritage language covers three groups of speakers who have ancestral links to a certain language: indigenous (first nation) language speakers, colonial settlers (e.g., Dutch speakers in the USA), and migrant language speakers like Arabic speakers in the host country. Thus, heritage language institutions are formed by authorities to maintain and support the languages in question; first because these languages 20

are not used by the pupils (i.e., younger members of the community) every day, and secondly as they are not taught within the ordinary language learning curriculum in schools. In such institutions, heritage languages are taught after the ordinary schooling. The term “heritage language” also has a negative side. García (2009: 60; 2005) points out that it puts focus on the past rather than on the here-and-now or the future, particularly in relation to linguistic practices. In García’s words, this term provides a way to ‘crack’ today’s homogeneous monolingual schooling of every different children in the United States, providing a space for the use of languages other than English in educating children. (García 2005: 602) The concept of heritage language is variously used (by means of other labels) in different language-educational contexts. Heritage language is most widespread in the North American academic and policy-making context. In Europeii and other parts of the world, other labels dominate. In the UK, it is often referred to as supplementary or complementary education; a term which puts the focus on educational initiatives organized outside of the English education sector. Initiatives are taken by minority groups themselves, and the curriculum often comprises both a community language and cultural issues (Blackledge & Creese 2010). In Denmark since the early 1970s, heritage language instruction has been termed “mother tongue education” (modersmålsundervisning) when compared to education in official minority languages (i.e., mindretalssprog) (Holmen & Jørgensen 1990; 1998). Terms such as native speaker, MT language user, heritage learner or speaker, and complementary or supplementary language learner all raise discussion. From an academic perspective, they have been used within applied linguistics to refer to speakers of first-acquired languages, speakers of other languages than English (in TESOL), speakers of migrant languages, and speakers related to a certain language by ancestral links.

21

Attitudes to these terms vary, though, a great deal: between the individuals and among group members who identify themselves with or are characterized by target languages of related institutions; between linguistic and ethnic majority language users and their relation to other languages spoken by linguistic minority groups; and in relation to authorities, policy-makers, and institutions. Linguistic majority speakers and institutions associate the speakers of migrant languages with a language, as speakers of their MT, regardless of their varied linguistic expertise and “‘levels of knowledge” (Blommaert & Backus 2011). Within an institutional context, MT classes refer only to linguistic minority groups, as if the mainstream language – Danish - is not a mother tongue (Kristjánsdóttir & Timm 2007). In short, language users use these terms explicitly and implicitly while positioning other speakers or themselves as linguistic out- or in-group members. Such conceptualization and the related attitudes reveal understandings of these terms. 2.2.3. MT in the Danish context In this section, I will sketch the context of MTE for linguistic minority pupils in Denmark, the national context of my study. The Danish Ministry of Education defines children eligible for MT tuition as bilinguals and as “children who have a maternal language other than Danish, and who do not learn Danish until they come in contact with the surrounding community or through the teaching of school” (UNI·C Statistics & Analysis for the Danish Ministry of Education: 48). According to Danmarks Statistik, 26.04 % of the population of Danish schoolchildren (grades 1 - 10) in Copenhagen (2013) were bilingual pupils in the official understanding of this term. Bilingual is used in different ways. One is understood as speakers with what is termed “native-like control of two languages”, and thereby with a focus on the speakers’ linguistic proficiency (Chin & Wigglesworth 2007). The concept of “bilingual pupils,” as defined by the Danish Ministry of Education, is not specified in relation to the language proficiency, competence, or which variety of the language is used. This may be problematic in practice in two ways: First, not all pupils are exposed to both languages at the same time. Second, when linguistic minority children enter Danish-speaking institutions and are exposed to 22

Danish, their first-acquired language will not necessarily develop equally with their Danish. In accordance with the European Union’s directive 77/486iii and following the Danish Education Ministry, the Danish state is obliged to offer mother tongue education in Danish public schools (folkeskoler) to a part of the bilingual pupils. The education is provided for free in addition to mainstream education to four types of what they term bilingual pupils: children identified with or originally from EU countries, from European Economic Area countries (EEA), from the Faroe Islands, and from Greenland.iv Following this directive (dated 22 July 2002), such pupils are offered courses in “the official languages of the mentioned countries” free of charge up until 6th grade. The Danish state’s attempts (or perhaps obligations) towards supporting the languages identified with these countries may indicate that Denmark is supposed to meet the educational needs of the labor migrants and their families from these countries. Also, it can be interpreted as an awareness of language diversity as a human right - the liberty of having, acquiring, and choosing a languagev in line with a value of Danish democracy. However, not every bilingual child in the educational system has linguistic roots in the mentioned countries. And this view misrecognizes the needs (if not rights) of other kinds of migrants who are associated with other countries than mentioned above, such as Arabic-speaking migrants. Thus, the Danish Teachers’ Association goes a step further with regard to the focus persons of MT courses and suggests that “all bilingual pupils should be offered mother tongue education regardless of their country of origin”vi. According to the Danish Ministry of Education, the pupils’ eligibility is based on one of two criteria: (1) the pupil or both of the pupil’s parents have migrated to Denmark from a country or region where the pupil’s MT is normally the language spoken (“det almindelige talesprog”), or (2) one of the pupil’s parents is born in Denmark, and the other (not Danish born) parent speaks the same language as the Danish-born parent. In this sense, MT is the language spoken in the pupil’s home. In the case of multiple home languages, the parents or the child must choose between them when selecting a MT class.vii In practice, these two criteria do not

23

cover all cases, though, as the study’s data show. Regarding the first criterion, the example is the case of families, where the common language of both parents - e.g., Kurdish - is different from the official language spoken in the parents’ country of origin or the language spoken at home or with family members - e.g., Turkish or with the family members in Turkey. As for the second case, the example is the case of families, where the common language of the parents - e.g., English - is neither of the parents’ first language - Farsi or Filipino - and the child having a good command of Danish is supposed to learn Farsi; which is a language she has no knowledge of. During the fieldwork I observed that some children did not speak the language in question at home, some enter the daycare system before they started speaking their respective home language, and some show various patterns of language use within everyday interactions. In fact, there is little discussion of the large differences within the group of such pupils, and thus, little discussion about their rather different needs. In the following section, I treat theoretical studies focusing on spatial distinction. This discussion will form the basis of my analysis of the data with regard to the binaries of home vs. school (or society), and country of origin vs. the country of residence. Departing from the theory of the territories of the self (Goffman 2010 ) and territoriality (Sack 1986; Storey 2001), I discuss how language use and understanding(s) of MT can be related to the understanding of space/place, as soon as the social actors (speakers) enter or are related to space. 2.3. Space, territory and language Language users construe meanings and express intentions through language, and all linguistic actions take place at a particular moment in time and in relation to a particular space. Space in relation to language users, language, and language use has been studied within sociolinguistics, psycholinguistics, and communication studies (see for example, Bakhtin 1981; Levinson 1996; Agha 2003). Drawing upon theories from human geography and social psychology, this section presents two types of relations between space and language users. By using these theories, I intend to investigate how language users relate themselves and 24

each other to certain spaces by means of an alleged MT. I also ask what type of space they establish through the so-called MT. In describing the theories, I follow a distinction - suggested by Mæhlum (2010) - between two usages of space as an overall concept. One is in the sense of the restricted and geographical location. I use place to refer to this sense. The other usage is in referring to interactional and social space, which is “appropriated and shaped” by social actors for the intended social activities (Blommaert et al. 2005); in using the term space, I refer to this view. I understand this in extension of Goffman’s (1964) definition of social situations in which social actors organize, ratify, and govern each other’s and their own actions. In this sense, the theories on space in relation to social actors and activities can be divided into two major groups. Thus in section 2.3.1 I review the theory of space proposed by Sack (1986), which is concerned with geographical and spatial territoriality, with a particular focus on real borders vs. conventional, predefined boundaries – I will clarify the distinction between the two concepts in section 2.3.1. In section 2.3.2., among theories on space concerning social interactions, I focus on Goffman’s (2010) theory of territories of the self. 2.3.1. Territoriality and territorial restrictions Individuals are physically situated in space: they act in space and claim space. They also exhibit relationships to spatial areas. The relationship between individuals and space involves the individuals’ behavior and attitudes within and with regard to space. Historically, one strand of thought regards the relationship between an individual and space as “innate” and “natural”, and another strand of thought considers this relationship as defined, conditioned, impacted, and even constructed by socio-political and cultural circumstances. This relationship is entrenched in power relations too, as a consequence of either controlling individuals’ behaviors within or excluding individuals from the space (Storey 2001: 9-13). The first strand is a biological perspective, which is not relevant to this study, and therefore I focus on the second strand in order to examine what type(s) of space is/are involved in the establishment and identification of MTE and MT.

25

Territoriality refers to the partitioning of space and asserting control over it. By gaining control of an area and of people’s activities, a group of people or an individual renders territoriality defensible. As Storey (2001) writes, this theory was basically built upon an ancient Greek paradigm of territorial units (i.e., city-states or polis), which were defined in relation to the three concepts selfsufficiency, security, and morality. By protecting goodness and morality, the ideal territorial unit could become or stay secure. Later critical views added to this understanding, claiming that the idealized view of the territorial unit was too simplified. Rather, one should focus on how various territorial units interact and provide economic opportunities, and therefore one should see them as politically organized (Storey 2001: 13-14). Drawing attention to the political reading of asserting territories within human geographical domain, Sack (1986) defines territoriality as a political and geographical strategy. Following this view, territoriality deals with “an attempt by individuals or groups to affect the interactions of others”. It involves controlling people or things by means of controlling a spatial area, which Sack names a territory (Sack 1986: 30-31). Sack’s approach enables us to include any types of territories, from geographical territories to socio-political spaces and forward to everyday relationships. This suggests that an individual or a group asserts control over a spatial area - namely, a territory - by establishing and maintaining restrictions over it; this influences other people’s access to the territory, and thereby controls the relationships within it. In short, territoriality concerns power relations (Sack 1986: 19). On the basis of Sack’s approach, we may say that territories are locations which are identified, delimited, and constrained by authorities. Thus territoriality happens in this order: territories are first identified and constructed by virtue of declared boundaries that an authority or authorities have established and the boundaries are maintained or administered by authorities or their agents. One concrete example is the geographical territory of a country defined by borders and international conventions. Boundaries are inherently related to the notion of transgression and thereby equal or unequal access to certain spatial areas. Individuals - whether dominant or not - within the territory should

26

respect the boundaries in terms of their activities or access to resources. In order to assert, maintain, and monitor territorialities, authorities introduce rules, norms, or prohibitions. This of course includes communication. Sack (1986) suggests that territoriality involves three key elements: First, classification by space, by which an area is defined as ours or off-limit to us; the classification is based on the objects’ “location in space” and not on their type; communication is the second key element. Through communication, we understand what is allowed within the restricted space and what is not. Norms are established and maintained within the territorial boundaries. Norms and boundaries regarding territories build upon conventions and symbolic units indicating direction, exclusion, possession, and so forth, with regard to access to territories and resources. I would rather say that norms and boundaries can be understood as semiotic signs that “carry semiotic value or significance to those who perceive them” (Agha 2007: 2). The third element in Sack’s approach is enforcing control over access to the restricted space, to objects, activities, or relationships within this space, or to the objects or relationships outside of it. As interactions within or with respect to the restricted space are influenced, controlled, and monitored by means of power, authorities and their agents might consider punishment in the case of transgressions. However, according to Sack, the power provided and reified by territoriality is not necessarily explicit and easy to perceive (Sack 1986: 20-34). My study aims to make sense of how individuals “spatial[ly] compartmentalize” (Sack 1986: 169) their language use as part of their everyday activity, parallel to other social interactions, and how they relate their “social spaces” (Lefebvre 2003) to the concrete places outside of their interactions’ here-and-now. Whereas Sack’s theory treats the mechanisms involved in territoriality of space, which deal with relations between individuals and their access to resources, other theories have to do with human interactions within restricted and territorialized space and their relation to space. In the following section, I briefly discuss Goffman’s 2010) territories of the self to provide a theoretical perspective on the relation between space and individuals as social actors.

27

2.3.2. Territories within interactional situations Goffman (2010), in the essay Territories of the Self, focuses on the social organization of individuals’ claim over a “field of things” which Goffman called a preserve. He also theorizes perceptions and expectations concerning distance from or proximity to preserves within social interactions. A claimant is defined as an individual or a part that makes a claim on possessing, controlling, or using a desired object or state; in Goffman’s terms a good, within a territory. A territory is a field marked by boundaries, and it includes visibly fixed personal space (e.g., houses), situational space temporarily perceived as possessed (e.g., a bench in a park), and ego-centric space (e.g., an individual’s surrounding space). By claiming and establishing rights over a territory, individuals can react against forces that violate the boundaries through a means or an act. Goffman names these counter-claimants. According to the theory, two oppositional claiming forces (and their agents) demonstrate their claim over and act for a desired state or object in interactional situations. I focus on two of the theory’s identified territories that I find most relevant in my study: The stall, and the possessional territory. The stall is a fixed temporary space, defined and identified by a claimant. The claimant can leave the stall for a period of time, like a reserved space at the beach that is signaled by a visible object. This space involves a multi-person possession. The possessional territory involves a set of objects and personal possessions that can be identified with an individual. In fact, objects one may own are considered and claimed as one´s possessions. Goffman asserts that claims to preserves are indexed or rendered visible by signs, called markers, and objects that indicate the boundary between two neighboring territories are termed boundary markers. As an example, a supermarket bar is used to separate customers’ shopping items. Markers are related to both claimant and claim, and keep intruders out; they demonstrate that it will be a violation to cross the boundary (Goffman 2010: 42). Words can also be markers and stop intruders to destabilize the claimants’ assertion and right to the preserve (Goffman 2007: 43). I find that Goffman’s territories of the self complement the theory

28

of territoriality on the level of everyday social interactions and humans’ placement within a specific space. Space is not necessarily concrete, but is constructed as such by means of objects, social acts and interactions, including language use. Soja writes that territories influence interactions among individuals and at the same time, they concern “components of self-identity” (Soja 1971: 33) as individuals constantly signal their relation to and control over territory and resources and the legitimization of their authority. To study how language use is defined, what type of language use is expected, and what types of social practices such language uses index during the construction of a social space established by MT education, I focus on registers of language. I describe this theoretically in the following section. 2.4. Language registers within geographical and interactional territories Individuals interact with each other through various forms of speech, which may be characterized metalinguistically as “proper language”, “standard language”, “dialect”, and so forth. Individuals acquire such speech forms through socialization throughout their life and intuitively describe their own and others’ speech forms, and thereby indicate to some extent the social awareness around these forms. In Agha’s (2006; 2007: 80) words, when being reflexively modeled, forms of speech become registers of language, and index an image of social stereotypes and relationships. According to Agha, “[s]emiotic practices differentiate a register’s forms from the rest of” what we traditionally call language (2007: 81). Registers can be understood as linguistic competences that are associated with social practices, people engaged in them, and people who can perceive, distinguish, and reproduce them in social interactions (Agha 2007). Registers are valued pragmatically, and within communicative practices both forms and values are transmitted among a population of language users (Agha 2007: 81). One example would be when a language user avoids using a linguistic register called “dialect” which the language user is otherwise identified with or has access to during interactions. The speaker’s effort to hide the dialect might be an attempt to align 29

him- or herself with other registers with higher social values in a particular situation, such as the so-called standard language. This may imply the asymmetry of social relations like the inter-group hegemony of the standard variety, and reflect other social and interactional dynamics, such as (the individual’s wish for) alignment with the group of interactants. Registers of language are formed over time and become subject to the acceptance, appraisal, or rejection of language users based on a system of valorization and countervalorization (Agha 2007: 24-26). In addition, a system of valorization about language use and perception is ideologically loaded (ibid: 29). An example is the case of the formation of and the consequent values around standard registers of languages as a result of their standardization - e.g., of Danish (Pedersen 2009) - and modernization and purism e.g., of Turkish and Farsi - respectively in Denmark, Turkey, and Iran. Standardization involves “the imposition of uniformity upon a class of objects” (Milroy 2001: 531). Similarly, in the process of standardization of a language, a variety of the language is understood as a separate entity from other varieties and is promoted by institutions as structurally invariant and uniform. In the case of Iran, the “standard” Farsi is sociopolitically and historically highly valued and has been associated with prevalent ideologies; it is treated as more appropriate to use in many social and official settings (see for example Perry 1985; Elling 2013). Geographically localized and standard registers of the respective groups of language users fall within this study’s interest in relation to the MT classes. Mæhlum (2010) describes a dialect as an idealized variety of language which is understood in contrast to the standard variety (which is idealized as well) and seen as specific to and rooted in a geographical area. Thus Istanbul dialect in Turkey is distinguishable from other Turkish varieties within the nation-state of Turkey. On the other hand, a standard register is a sociopolitical norm and a baseline against which other geographically localized varieties - and, in general, other perceived registers of the language - are “measured” in Agha’s words (2006: 24).viii This language variety is often the prestigious variety among the social elitesix and linked to authorities such as those controlling

30

educational institutions (Weber & Horner 2012: 17). In Mæhlum’s words, the standard register is historically tied to a geographical region - often the capital of a nation-state. Over time, the sociopolitical centers have ideologically selected this variety as the standard, natural, and beyond-discussion form of the language. This in turn impacts other regional varieties in an “asymmetrical social relation,” and generates ideological associations around different varieties: labels attributed to the dialectal registers may be considered less prestigious, informal, oldfashioned, or rural, while the standard variety is seen as prestigious, urban, modern, or formal (Mæhlum 2010: 21-24). In the present study language purism is of importance as a type of language ideology. Language purism appears as an outcome of social and historical changes within nation states. Language purism is the consequence of an idealized discourse and can be situated within the project of modernity of nation states. Authorities and language users wish for a language cleansed from unwanted linguistic elements and thereby language users in a social context affected by language purism may police each other’s language use (Preston 2005; Albury 2016). Studies show that language purism can emerge from the issue of language contacts e.g., in the case of attempts to cleanse a language from foreign lexicon, grammar, and pronunciation. Likewise, it may be related to all elements of other language registers and varieties than the valued standard variety, such as sociolects and dialects (Absillis & Jaspers 2016: 1). Language purists ignore the fact that language is dynamic and constantly subject to change (Weber & Horner 2012: 20). In this sense, the language is treated as an entity that should be safeguarded from an imagined contamination by unwanted elements and/or should be purified. According to Langer and Davies (2005), when language purism receives an “official” reason particular words or constructions are discriminated and named as “illogical” or “contaminating” features. A second reason for language purism might be the fear that the standard language is threatened by linguistic features associated with foreign or “lowerclass” cultures (Langer & Davies 2005: 4-6). I return to this ideology with regard to the socio-historical background of the languages in my study.

31

Valorization and counter-valorization of ways of speaking and their consequences will be demonstrated analytically in this study, as I explore what language registers are used and suggested in MT classes. Moreover, I will explore how they are expected to be learned by the new generations and used by them as well as by the adult participants, that is, parents and teachers. Is the standard register spoken or preferred in the respective language groups, and standard in what sense? Are other varieties allowed? And what type of varieties? Also, I investigate whether and how the alleged MT language users relate the standard or local registers to geographical and social territories. How they connect one variety rather than another with reference to their countries of origin and to the Danish borders? And why and where do they do so? I address these questions through the second part of my data analysis.

32

3. SOCIO-CULTURAL AND LANGUAGE-IDEOLOGICAL BACKGROUNDS OF TURKISH, IRANIAN AND ARAB MIGRANTS In this section, I provide demographic information regarding the three groups of migrants the present study focuses on (see part 4). With this information I aim at presenting the historical reasons for these populations residing in Denmark. Secondly, I present the socio-political background of certain linguistic ideologies with which these groups of migrants identify their languages of origin. Based on this knowledge, I aim to explore what understandings are formed in relation to the focus groups’ languages and countries of origin. In the case of Arabic-speaking immigrants and the pupils of the Arabic MT classes, I follow the institutional label of speakers with Arabic as their or their parents’ first language. Despite their diversity based on the countries of origin, Arabic speaking migrants are regarded as connected to a particular linguistic community of Arabs, both from the Danish institutional and in-group perspective. Institutionally, they are distinguished from other linguistic minority groups such as Turks. From an ingroup perspective, the label “Arab”, in relation to the target language of the MT classes, refers to various representatives of Arabic countries with certain associations: Arabs as Muslims, Arabs’ common culture and lore, and Arabs in relation to the Arab World. Likewise, Arabic is an overall term corresponding to a linguistic and communicative medium, as well as an educational medium at the written and literacy level. Thus, Arabic is also regarded as a key to entering and communicating with and within the Arab World. This is to be distinguished from the diverse varieties of Arabic as oral, communicative medium within the national borders. People in the territory of the Arab World are educated in one (or two) variety(/ies) of Arabic, namely Standard or modern Arabic, which differ from the regional and local Arabic dialects. One can say that the Standard is an idealized register of Arabic which by means of linguistic uniformity attempts to render the Arab World to a linguistic unity (Haeri 2000).

33

3.1. Demography of the migrant groups with Turkish, Iranian, and Arabic backgrounds The Turkish migrants form one of the largest migrant groups in Western Europe, with reasons of migration including employment, family reunification, and political conflicts in Turkey. This group’s migration started in the late 1960s (Akinci 2008; Backus 2012). Approximately 62.000 Turks and their children live in Denmark (2016), being the largest migrant groupx. 1.31 % of the population of Copenhagen has a Turkish background (7.732 people). Among these, several have a Kurdish-Turkish background. It is, however, difficult to provide exact statistics of this group and the population remains a rough estimate; the reasons are beyond the scope of this publication. According to studies from different Western countries, apart from speaking the societal majority language, Turkish migrants maintain a strong transnational connection to their families in Turkey (Böcker 2000) and thereby create transnational identity. The case of migrated Iranians differs strikingly from the Turkish case. Migration of Iranians to Western countries mainly started in the late 1970s and mid-1980s due to political change and the Iran-Iraq war (Wright 2010). Migration from Iran still continues, following Iran’s economic and political instability. Denmark Statistics counts 19.242 migrants with Iranian background in total. 3.668 Iranians reside in the municipality of Copenhagen (Statistikbanken 2016). According to The World Bank, 369.8 million people live in the Arabic territories, spread over 22 countries of the Arab World. Immigrants from Iraq and Lebanon have been the most populous Arab groups whose migration to Denmark started in the 1960s and 1970s.xi These two groups contain 31.322 Iraqis and 25.907 Lebanese in the whole country. Approximately 6.678 Iraqis and 4.896 Lebanese live in Copenhagen. The migrants who are originally from Palestine relate statistically to a former stateless population or with Lebanon as the country they come from; however, this group forms a rough estimate. Two other large groups of Arab migrants originate from Morocco (5.194) and Syria (1.212). People who are originally from these four nations fall within the scope of this study, as the participants in this study have one of these national origins. 34

3.2. Socio-cultural background of the three languages: Arabic, Farsi, and Turkish In this section, I elucidate the issue of language standardization of Arabic, Farsi, and Turkish, as a part of the language history of related territories. I briefly present ideologies pertaining to the standardization of language, understood as an isolated entity, “pure languages” in Iran and Turkey, and standard language versus regional varieties within the Arab World. Language ideology and purism regarding Farsi and Turkish: An interesting fact concerning the Turkish and Iranian migrant groups is the tradition of language purism in their countries of origin. This is a relevant issue for this thesis with a focus on the relation between language, language users, and countries of origin. As mentioned in 2.4, linguistic purism is concerned with a population of speakers’ attempts to cleanse a language from features imagined as undesirable and foreign, e.g., foreign words and loan words (Thomas 1991). This ideology also deals with discussions of “standard and non-standard varieties [of a language], preserving older varieties and rejecting younger ones, [and] the role of language in nationalist ideology” (Langer & Davies 2005: 2). Language speakers, in this sense, have concerns about their language and perceived dangers such as modernism, foreign languages, and non-official and geographically localized varieties. Speakers may try to take certain measures to prevent or act adversely to the undesired developments on both an individual and institutional level. Linguistic purism is tied to the historical background of Turkish and Farsi and the modernization and secularization projects of the two countries (separately) during the 1920s (Perry 1985; Paul 2010). In Turkey, the government founded the Turkish Language Association (Türk Dil Kurumu) in 1932 to invest in Turkish and to cleanse the language from Arabic and Persian linguistic features, which were regarded as traditional and dated (Jernudd 1989). The language reform was initiated by introducing the Latin alphabet and new spellings of Turkish words in 1928. In Iran, the first Language Academy was established in the early 1930s – again, with a focus on finding proper equivalents and neologisms for 35

Arabic and foreign words based on a historical and linguistically pure Farsi (Paul 2010). Language politics, as a part of modernization projects, in both countries deal with separate norms. In Iran, Farsi (with Tehran’s dialect in focus) has for decades been considered the standard, national, and official language among different local languages, dialects, and regional varieties (Perry 1985). In Turkey, Turkish has been the only national and official language, compared to other minority languages like Kurdish. Turkish has been standardized for decades by downgrading local dialects or varieties, too; this fact is seen in everyday perceptions of people with non-standard dialects, particularly with Eastern Turkish traits (Demirci & Kleiner 1999). Likewise, such ideologies can be noticed in the fact that most media programs in Turkey are produced in standard Turkish. In a personal conversation, a Turkish anthropologist states, TV Turkish is standard Turkish (…) I think Turkey was a rather successful example in attempts to standardize the language nationwide, when people's native language is Turkish. There aren't really dialects but accents, about pronouncing every letter. Language ideology and the identification of Standard Arabic: Historically, two major varieties of Arabic have co-existed since 67th century A.D.: the prestigious high variety (lugha al-fusha or “pure language”) and the colloquial, low variety (lugha al-ammiya or “colloquial language”) (Dakwar 2005). While the first one is associated with written Arabic, religious texts, political and literary discourse, and formal interactional settings, the latter group is connected to everyday oral communication. The high variety, or Fusha, is also historically associated with the language of Islam (both as a religion and a civilization), with classical texts, the Arab identity, pan-Arabism, and political resistance to colonialism of countries such as France, England, and Italy (Haeri 2000). The Ammiya group extends over a large geography within the Near and Middle East and North Africa, and although it has a myriad of subvarieties depending on the geographical regions, it makes a continuum of huge diversity with lexical, grammatical, and

36

phonetic differences (Suleiman 2003). Ammyia, in general, can be related to two major groups in terms of their shared heritage: the Machrek (Eastern) group and Maghrebi (Western) group. Machrek countries consist of Syria, Lebanon, Jordan, and Palestine (i.e., related to the Levantine Arabic); Egypt and Sudan; Iraq and Saudi Arabia; and the Persian Gulf states. Maghrebi countries consist of Mauritania, Morocco, Algeria, Tunisia and Libya (Maamouri 1998). For decades now, the impact of the linguistic difference of the two major varieties - Fusha and Ammiya - has motivated debates concerning acquiring Arabic in educational settings and literacy among children in the Arab World (Dakwar 2005; Maamouri 1998). In the field of education in the Arab world, Fusha is the baseline and the standard medium for literacy purposes and, over centuries, has been transmitted across generations. But while it is the language of education and literacy, nobody has it as their first learned linguistic variety, their “mother tongue” (Maamouri 1998: 33). In other words, the linguistic difference between Fusha and Ammiya varieties has resulted in difficulties for many children in learning the codified Arabic (Dakwar 2005). Yet, to eradicate this linguistic distance and its potential consequences, intellectuals and reformers in different countries of the Arab World have suggested language politics. Prime examples include a renovation of Fusha (Haeri 2000) or a promotion of a variety of Ammiya on the basis of Fusha that might trigger a new intermediate variety (Maamouri 1998). It is also noteworthy that the intra-lingual code-switch (i.e., inserting local varieties inside Fusha) has been historically subjected to debates and even suspicions.xii As for migrant families with various countries of origin - like the case of my study - I think that the probable use of Arabic TV channels from various countries may provide their next generation with programs in both Fusha and at least a hybrid of Fusha-Ammiya. Though, I do not have further information about the latter form of language practice. To sum up, I have provided an overview of the historical background of certain linguistic political measures and debates from the language communities to which the participants in my study are related. These debates are not only remote memories that the Turkish, Iranian, and Arab migrants might have, but they also continue online (via social media) as well as across and within

37

the borders of the countries these languages are related to. I use this historical fundament to analyze and discuss the data in question in this study.

38

4. METHOD AND METHODOLOGY This chapter presents an overview of the methods and methodology of the study. The study is based on a larger empirical project that investigates the issue of Mother Tongue Education (henceforth, MTE) in Copenhagen. The overall project of MTE for Linguistic Minority Pupils in the Super-diverse Metropolis of Copenhagen (Super-MOTE) aims to provide insights on language and literacy practices within, understandings of, and attitudes towards MTE. The overall project includes four MTE settings: Arabic, Farsi, Polish, and Turkish. However, the present study restricts its focus to three language settings – namely, Arabic, Farsi, and Turkish. This restriction is due to three reasons. First, these languages have been in contact as a consequence of historical events in particular: (1) the spread of Islam (since 613 A.D.), (2) the Mongol invasion of Iran (1219-1221 A.D.), and (3) the expansion of the Ottoman Empire (1453-1606 A.D.). While the first and the third events caused the contact of speakers and thereby the three languages, due to the second event, Mongolian as a Turkic language influenced Farsi, and Farsi in turn influenced the Turkic language in the Western parts of Iran (Perry 1985; Doerfer 1981). In the modern history of the three languages since the 1920’s, linguistic attitudes with a focus on “purism” have developed and circulated among Turkish, Farsi, and Arabic language users and officials towards local minority languages and vernaculars, towards Arabic and Farsi in the case of Turkish language users, and towards Arabic and somewhat towards Turkish in the case of Farsi language users (3.2.). This introduces the second reason for me concentrating on the three settings: language users display relatively different attitudes towards the cultures (despite many cultural similarities) and the official languages of three geographical areas. Such representations circulate among the communities associated with Farsi, Arabic, and Turkish. Finally, being an Iranian ex-patriot, my general knowledge is limited to these languages and their related cultures, whereas Polish is beyond the scope of the present study. Methodologically, the study of the pedagogical settings of these three MTE classrooms and the attitudes of the studied 39

participants towards MTE builds on Linguistic Ethnography (Blackledge & Creese 2010; Rampton et al. 2015). In what follows, I provide a detailed account of the methodology, data collection and analysis, and participants. 4.1. Fieldwork study Drawing on a combination of linguistic and interactional analysis, along with longitudinal ethnographic analysis, linguistic ethnography is a theoretical and methodological framework (Maybin & Tusting 2011; Rampton et al. 2015). As Creese writes, linguistic ethnography is the investigation of “the interplay between language and the social, the patterned and the dynamic nature of this interplay and the processual nature of meaningcreation in the making of context” (2008: 229). During ethnographic fieldwork within a selected field, with one or multiple sites, as a point of departure the fieldworker conducts descriptive observations based on broad questions. Once the fieldworker gets familiar with the site(s), she or he can distinguish characteristic traits from each other and from the less common characteristics. The next step is to narrow down into “more focused observations (…) of specific portions of the people, behaviors, times, spaces, feelings, structures, and/or processes” (Adler & Adler 1994: 381). Following this method, the fieldworker aims to explore reflexively the use of language of participants in various social encounters within a specific time and space frame. In this sense participants’ language use is attached to two additional sociolinguistic properties, and these properties all together construct what is termed as the total linguistic fact (Silverstein 1985). A total linguistic fact is composed of linguistic form, language use, and the indexical meanings with reference to which the language user conveys intentions and ideological biases. Linguistic ethnography is historically rooted in the tradition of ethnography of communication which aimed to investigate language use in connection with language users’ perspectives within social situations (Maybin & Tusting 2011). Linguistic ethnography departs from participants’ perspective of their habitual life and their social context and - on the basis of language data - studies their reflections and understandings around the field about language, actions, and 40

social relations. Fieldworkers collect empirical data of different types and kind through the documentation of observations, immediate experience of the field, and the field participants, particularly with regard to the participants’ everyday practice (Maegaard & Quist 2005). Fieldworkers collect field notes, interviews, recordings of the encounters, fieldworkers’ personal communication with the participants, and pictures taken during the fieldwork, and similar methods. The multiplicity of data improves the study in terms of validity and makes it possible to reach various understandings. Team ethnography is suggested in order to gain higher validity and enrich the understandings of social dynamics existing in the field. In a traditional anthropological fieldwork the fieldworker was considered a “lone ranger” (Creese et al. 2008: 199), whereas in team ethnography a team of fieldworkers may be present in the field and observe and discuss the field from various perspectives. In this sense, team ethnography, suggested by Erickson & Stull (1998: 18-20), provides the opportunity to study the participants with a wider perspective from different vantage points, and often involves researchers of various social, cultural, and ethnic backgrounds. In this tradition, fieldwork is conducted in different settings (Austin 2003). However, a disadvantage of this type of ethnography is that the presence of many fieldworkers at a time may impact the organization of interactions and even the participants who might find themselves observed. This concerns the overall project of the present study. In the overall project, Super-MOTE, a group of team members (all affiliated with the University of Copenhagen) has been conducting fieldwork since 2013. Structurally, the researchers have been responsible for four sub-projects, individually or in groups of two. In each sub-project, there has been at least one fieldworker with proficiency in the target language of the MTE classroom. Team researchers have discussed their reflections about observations within the sites; such discussions have a number of purposes: to gain a more comprehensive sense of the topic; to increase the validity of the study; to gain and provide comparative and contrastive insights; to clarify vague issues from their experiences, particularly regarding cultural and linguistic differences; and to exchange ideas about ongoing publications. Discussions have usually been monthly.

41

4.2. Fieldwork sites The overall project’s leader and the researchers affiliated with the Arabic and Turkish sub-projects selected the MTE classes which took place in a linguistically and ethnically diverse public school in southeastern Copenhagen. The Super-MOTE project overlapped with another linguistic ethnographic project, namely the Copenhagen Studies in Everyday Languaging, also based at the University of Copenhagen. A number of researchers in the SuperMOTE project have been involved in Copenhagen Studies in Everyday Languaging and a few children were focus persons in both projects. This overlap brought about a favorable result in two ways: (1) to achieve better understandings concerning the research questions, because of the already existing insights; (2) to assist the Copenhagen Studies in Everyday Languaging in expanding knowledge about aspects of the pupils’ language use in other settings than previously recorded. MTE classes took place in the continuation of the ordinary school schedule. The sub-projects overlapped in terms of various topics, such as the relationships of mother tongues and geographical and social spaces, the private and public organization of classes, the socio-cultural differences of participants with minority linguistic backgrounds, and so forth. Such thematic similarities and nuances assisted the fieldworkers in achieving a wider knowledge of MTE. In the Arabic MTE case, there were two different classes twice a week (1 h. 10 min. each). One class was devoted to a younger group, while the other to an older one. The classes always took place in one of the regular classrooms on the second floor of one of the buildings specified for the primary program. This class was referred to as the “Arabic classroom”. The walls in the classroom were covered with Arabic posters (published in Egypt), upon which illustrations of animals and parts of the body were accompanied with lexicon in Arabic and English. I visited this classroom once to conduct fieldwork in the Turkish site. Arabicconnected artifacts inside the classroom and Arabic books in pupils’ drawers indicated that the room was reserved for an Arabic language class. The Turkish class was offered once a week (1 h. 15 min.). The class took place in two different sites during the fieldworkers’ 42

visits: In the fall semester, the Turkish class was located in the Arabic classroom. After January 2014, the class moved to another room located on the top floor and usually employed for special education. There were no Turkish posters or artifacts associated with Turkish in either of the rooms. Both the Turkish and Arabic classes were publicly arranged and funded. The two Farsi classes in which I worked as a volunteer took place at different places. They also differed from the Turkish and Arabic MTE classes as they were privately organized. The first class was situated in a culture center inside Copenhagen with international perspectives. The second was located in a public school in the suburb of Copenhagen. The teacher (anonymized as Mansour) and the assistant teacher (I, NG) taught in both classes. The class inside the city took place every Saturday, and the class outside the city took place every Friday (each lasting 2h. 30 min.). The two classrooms lacked any artifacts associated with Farsi or Iranian culture. The first classroom had a whiteboard that was rarely used. The walls in the classroom were covered with pictures that had no obvious relation to language instruction. An excerpt taken from a diary written by a fieldworker displays this idea: De billeder der hænger på væggene illustrerer en diverse skare af kvinder fra de forskellige arrangementer, som Kvindeforummet har holdt. [The pictures hanging on the wall illustrate a diverse flock of women from different events that “Women’s Forum” has organized.”] (Field note by XLN; 07-09-2013; my translation) This excerpt indicates the fieldworker Lamies Nassri, (XLN) is puzzled by the incongruence of the class content - being Farsi instruction - and the class environment, i.e., pictures displaying the Women’s Forum’s activities. In addition, the objects from the classrooms were never used pedagogically during the MT instructions. However, objects inside the classroom may occasionally have motivated conversations among the pupils in different sites.

43

In the following section, I detail the participants in the study, which will be followed by an overview of the data achieved within the sites. 4.3. Participants and data Several types of participants were observed in the fieldwork sites: MT teachers, MT pupils, and the pupils’ parents and (in a few cases) grandparents. This section presents the social background of the participants, followed by an overview of the data collected among these participants. 4.3.1. Participants MT teachers MT teachers were the project’s first contacts. In order to visit and participate in the MT classes, fieldworkers asked the MT teachers for permission. In the case of the Farsi MTE class, I had asked for permission prior to the start of 2013. In fact, I had been working as an assistant teacher since September 2011, and I already knew the participants to some extent. The presentation to the Arabic teacher (anonymized “Aslan”) took place in February 2013 by three researchers. As for the Turkish part, the Copenhagen Studies in Everyday Languaging project was presented to the Turkish MT teacher (anonymized “Mehmet”) before the beginning of SuperMOTE by Professor Jens Normann Jørgensenxiii. When Super-MOTE was launched, Prof. Jørgensen introduced him to the project leader (MSK). However, as Jens Normann fell ill, fieldwork in the Turkish classes was postponed until the fall of 2013, when the Turkish student (XON) joined the project team. Several of the fieldworkers knew the Arabic teacher prior to the project, as they had conducted other ethnographic research at the school where these two MTE classes took place. The teacher was therefore also somewhat familiar with fieldwork and the presence of fieldworkers at the school. The Iranian teacher knew almost nothing about methods of such studies, nor did the parents or the pupils and their unfamiliarity with the methods triggered a challenge for the fieldwork, particularly in regard to students’ dis-preference to be audio- and video-recorded, to be interviewed, and to respond to 44

questionnaires. And finally, although the Farsi MT teacher on a few occasions encouraged some parents to be interviewed, he also showed remarkable consideration for parents’ dis-preference for contribution (See Karrebæk & Ghandchi, forthcoming). According to the interviews with the MT teachers, they had very similar teaching backgrounds and experiences from their countries of origin Iran, Iraq, and Turkey. They were primary school teachers prior to their migration. All were bilinguals with Kurdish as their first language and the official language of their countries of origin (Farsi, Turkish, and Arabic) as their second language. In Denmark, the Arabic and Turkish teachers had pursued pedagogical education at Teachers’ College. Mehmet (Tr.) and Mansour (Fa.) had started teaching MTE in the late 1980s, and Aslan began teaching MTE in 2002. Since then, Aslan and Mehmet had been working in the public schools, while Mansour had switched to private institutions. This year was mentioned in their interviews as a turning point in the history of MTE in Denmark. In 2002, the then government stopped nationally funding MTE to all bilingual pupils. This led to municipalities’ closing many of the classes - except for the municipality of Copenhagen which continued to support MTE. Since 2015-2016 MTE has continued to be offered in the municipality of Copenhagen without any charge to school children between 0th-5th, and with charge to students between 6th-9th grades. This offer was decided irrespective of children’s EU or non-EU origins. However, this has not been the case for the children outside of the municipality of Copenhagen, and either families or the municipalities have to pay for MTE if interested. Since 2002 and right after the shortages of the national funding of MTE, Mansour has continued his work as a private teacher in 5 MTE classes, two of which I followed. Aslan taught Arabic with an assistant teacher, anonymized as “Noor”. She was originally from Lebanon and had no education from Denmark or Lebanon. As for pedagogical methods and materials, the three teachers were free to choose. All MT teachers are in general advised by the directive of the Danish Ministry of Education (Undervisningsministeriet 2009: 23) to choose the appropriate materials. Although the Turkish class had textbooks, Mehmet

45

preferred to primarily use a Smart board, the internet, and certain texts of children literature produced in Turkey. In the case of Arabic, Aslan had chosen a set of teaching materials (in three levels of 0-2-3) published in France. These materials were specifically produced for children who grow up outside the Arabicspeaking countries. Mansour used two sets of books produced in Sweden in 1996-97. While one series met the primary level of Farsi language learning in Iran (though with somewhat different topics and illustrations), another one was suited to more advanced pupils. Texts in the two series were primarily based on literature and were occasionally illustrated. As most of the Farsi pupils belonged to the younger or less proficient group, they only rarely used the second set (2 occasions in one year). Like Mehmet, Aslan incorporated digital media available in the classroom - though not as frequently as him. Mansour, however, was not interested in using materials other than books and, later, copies of exercises as supplementary materials. In order to inform the families of the fieldwork and receive their consent for recording the classes, the project folders were delivered to the families through the teachers or by the project team’s direct contact with the families. Pupils and their families In general, the pupils in study can be divided into two groups: younger and older children. The younger group of Arabic classes ranged from 1st grade to 3rd grade, and the older group comprised 4th and 5th graders. The younger group of Farsi classes was composed of 5 to 10 year old children, and the older group of 11 to 17 year old children. The Turkish class was composed only of young pupils from 1st through 4th grade. Turkish was offered in one class only; Arabic instruction was divided into two classes. The Farsi classes differed exclusively. The urban class included one group only, and the suburban class included two groups. In the latter class, the two groups were distinguished by means of their break time and the place they sat (or were placed) together. All three sites included siblings. In Farsi classes, this was exploited to legitimize certain mobility between the two groups of younger and older students, such as when the younger children preferred to sit beside their siblings or when they moved to ask a question related to their tasks. 46

It is noticeable that in Farsi classes, unlike the two other classes, the children in both age categories differed in a range of relevant characteristics. First, they had different levels of proficiency in the target language, the language of their “country of origin” or “heritage country”. This includes both comprehension and production, as some of them communicated fluently, while some produced few words and some made almost no attempts to communicate in Farsi. In terms of motivation, a diversity of pupils was observed. Some of them showed motivation to learn or practice the target language, while some lacked motivation or interest, and demonstrated this in various ways. For example, these demotivated pupils were not interested in pursuing the pedagogical practices; they frequently asked what time the class finished, when they could have a break and at what time their parents would pick them up. They were noisy or played around, and did not focus on their exercises. On many occasions, they stated that the classes were boring. However, factors outside of simple demotivation might have been in play as well. The MTE classes took place after the ordinary schedule of the schools or on weekends and MTE is generally not valued or recognized among children, or in the Danish educational system. Such factors are beyond the scope of this work to explore, though. The children differed in terms of their family backgrounds. Some were from families with mixed-marriages (parents with different linguistic and nationality backgrounds), while other children’s parents came from a single country and spoke the same language. In the Arabic classes, there were children with four major Arabic national backgrounds: Iraqi, Moroccan, Palestinian, and Lebanese. In the Turkish class, some children had KurdishTurkish origins and one child with Uzbek origin. In the Farsi classes, the parents represented various national origins: Afghanistan, Denmark, England, France, Germany, and Iran. To the researchers’ knowledge, this variation resulted in a language preference strategy at home: in many cases, Danish or the language of one of their parents was the dominant language. This aspect had not totally excluded other potential languages at home, where Danish or another familial language, such as Kurdish, was often spoken as well.

47

In addition to this diversity, families also varied in terms of their migration histories, concerns, and motivations. Some of the families belonged to second- or third-generation exiled Iranians. But I am not aware of the details in all cases and I will not treat it any further. Yet, I believe that different migration histories had consequences, at least in the Farsi case. I faced instances in which children (as the third generation) were not exposed to the class target language very often. Instead, their grandparents insisted on presenting the heritage culture to the children in the form of heritage symbols and celebrating heritage ceremonies. I failed to obtain home recordings from the families, and questionnaires that I distributed raised discussions and suspicion among families and the MT teacher. Many Iranian parents wished a class free of politics and ideologies for the children to acquire the target language, and literacy within an ideology-neutral space meant a lot to them. Many parents had direct or indirect involvement with politics as their reasons for migration and that possibly led to feelings of insecurity and mistrust. This attitude that was presumably threatened by the ethnographic methods of recording or observation added to concerns in different phases of my fieldwork (Karrebæk & Ghandchi, forthcoming). 4.3.2. Data, devices, and data process The data collected from the three sites are composed of field notes based on observations and fieldworkers’ reflections; audiorecordings for Arabic, Farsi and Turkish MT classes; videorecordings for Arabic and Turkish classes; a few short selfrecordings, mainly from the pupils of the Arabic classes; group conversations; shorter conversations with MT teachers (and parents in the case of Farsi classes); interviews with parents and MT teachers; and questionnaires for Farsi classes. Table (1) demonstrates a brief overview of some of these data:

48

MTE classes Field notes Interviews with parents Interviews with teachers

Arabic (two age groups) 47

Farsi (two sites)

Turkish

58

21

8

16

3

3

1 (2 parents present) 1

Table 1. Overview of data Parts of data from the three sites overlap. By this I refer to cases in which both audio- and video-recordings contain the same sound files; different audio-recorders have covered different parts of the classrooms; and shorter conversations with the MT teachers and parents are recorded parallel to a class recording. All of these indicate that a precise amount of the data collected over the whole year of fieldwork is barely possible. Audio-data in the overall project have been recorded on Mp3 digital voice recorders. In order to capture a wider picture of the sites in two sites of Arabic and Turkish classes, video cameras were used occasionally. Those responsible for the Arabic and Turkish sub-projects asked the families to declare their consent for the project-related enquiries and recordings. In the case of Farsi classes, I informed the main teacher and the families prior to audio recording. Accordingly, they agreed to audio recording, but video recording was a matter of polite hesitations or refusals. The process of children getting acquainted with the recorders was also a long one – right up until the last session of my fieldwork, some asked me why I recorded in the class. All forms of data have been saved on a secure server, hosted by the Centre for Language Change in Real Time (LANCHART) and based at the University of Copenhagen. The participants are anonymized. I have selected a number of interactional sequences and excerpts of field notes taken from the data collected in the three sites to respond to my research questions. Regarding transcriptions and conventions, the researchers who were in charge of the Turkish and Arabic sub-projects transcribed data using the software CLANxiv. For the Arabic and Farsi data -

49

recorded in the Perso-Arabic alphabet - a Romanization of the transcriptions was applied instead of providing data in original, gloss, and translation characters. The two languages do not follow the same Romanization system as they are based on two different phonetic systems. I employed Uni-Pers Romanization system for Farsi.xv

50

5. ANALYSES 5.1. Analysis (I) 5.1.1. Whose mother tongue wherein? This section which falls in two parts is related to the two first strands of theories concerning revisiting the terms MTs and MT users, and concerning space and place. The first part provides three sets of examples in relation to the respective language groups. The examples illustrate various participants’ understandings of MTs, MT pupils or users, and preferences about language use, as well as the pupils’ knowledge of and communication in the languages as used at home and in MT classes. I examine these issues mainly through the ethnographic data from the adult participants’ reports of the pupils and from class interactions. The second part of the section concerns different understandings of the relation of the MT pupils’ use of MTs to the places which these languages are tied to normatively. Through another three sets of data, I examine how adult participants and pupils negotiate their understandings about the potential linkages of the MT to geographical places. Throughout this chapter, I use the terms I introduced in chapter 2 and do not spell out the theoretical reasons for and implications of my choice. “She cannot say it in Turkish” The overall ethnographic research observations (including my own) noticed different specific patterns of language use related to the MT classes by varied participants connected to these sites. The participants commented meta-discursively on these patterns in their various interactions with the fieldworkers. The participants’ value-laden accounts reflected what they understood by using and learning a “language” referred to as MT and Danish, and where and how the participants preferred these languages. To show this, I start with two field note extracts written by two researchers with two years of interval in the Turkish class taught by the same teacher. One is written by Jens Normann Jørgensen (JNJ), and it gives a background image of some of the Turkish MT pupils’ language use before the project Super-MOTE was launched. At a 51

first glance, the extract demonstrates that JNJ had a particular focus on codeswitching between Turkish and Danish, but also on the use of Turkish during the sessions: Example (1a): Speaking Danish to the teacher Fieldnote by JNJ; 23-8-2011. Özlem og Fadime fortsætter med at tale dansk med hinanden (…) Mehmet henvender sig til dem på tyrkisk, og for det meste taler de også tyrkisk til ham, men ikke hele tiden. Selv når de taler dansk til ham, svarer han på tyrkisk. Mehmet nævner hurtigt de tyrkiske betegnelser ((ord om år, årstider og ugedage; my comment)). Det ser ud til, at han tror, at pigerne kender dem ((senere viser det sig ikke at være tilfældet …)) [Özlem and Fadime continue speaking Danish together (…) Mehmet addresses them in Turkish. And generally they also speak Turkish to him, but not always. Even when they speak Danish to him, he answers in Turkish. Mehmet names quickly the Turkish characters ((e.g., words for year, seasons, and weekdays; my comment)). It seems that he believes that the girls know them ((it becomes apparent later that it wasn’t the case …))] JNJ reports of the constant interaction of two Turkish MT pupils (the girls Özlem and Fadime) in Danish. When the pupils initiate a dialogue in Danish, the teacher (Mehmet) replies to them in Turkish, the same language is used when he addresses them. In response to Mehmet, the pupils use either Turkish or Danish interchangeably. JNJ’s observation can be summarized as follows: 1)

Danish: Interaction Pupil 1  Pupil 2; Initiating communication Pupils  Teacher

2)

Turkish: Interaction Teacher  Pupils; Response Teacher  Pupils

3)

Danish/Turkish: Response Pupils  Teacher

52

What these schematized patterns of language use show is that the only participant, who insists on using Turkish without shifting to Danish, is the teacher. He also seems to have certain assumptions concerning the pupils’ Turkish command. In the second part of the observation, the teacher quickly names some Turkish words. The observer notes that this must be because, Mehmet has taken the girls’ acquaintance with the words for granted; thus he does not introduce them. Yet, according to JNJ’s observation the assumption was not correct: not only do the pupils not speak Turkish together, but they often shift between the two languages and do “not always” speak Turkish while addressing the teacher. This remark adds to the MT teacher’s assumption about MT pupils – namely, that the MT teacher prefers or perhaps presupposes the pupils to use and understand Turkish inside the MT classroom while the reverse is reported. A similar pattern of language use inside the Turkish class is observed by another researcher, Martha Sif Karrebæk (MSK) in a field note two years later (2013): Example (1b): “He does not understand” Fieldnote by MSK; 14-11-2013. Når børnene ikke taler til Mehmet, taler de dansk. Men hvis de henvender sig til ham på dansk, siger han, at han ikke forstår. [When the children do not speak to Mehmet, they speak Danish. But when they address him in Danish, he says he does not understand.] MSK’s observation points again to a monolingual pattern of Turkish use (i.e., the teacher speaks only Turkish with children), despite the teacher’s insistence on receiving pupils’ response in Turkish. In fact, the teacher refuses to answer in Danish and even claims that he does not understand Danish, presumably to encourage them to re-formulate their questions in Turkish. He does not provide an explicit metalinguistic model, such as say it/speak in Turkish. Fieldworkers did observe this in other instances. But the claim of not understanding Danish was one of Mehmet’s returning pedagogical strategies. Likewise, it contrasts with Mehmet’s knowledge of Danish, as noticed by other fieldworkers through conversations. Mehmet uses this claim of not 53

understanding Danish to inform the pupils of the teacher’s preferred language, and there is a need for that, as it differs from the pupils’ preference. Additionally, though, it shows that the communication between the teacher and the pupils is not aligned, as the teacher insists on normative aspects related to the language use and the pupils on the content and flow in interaction. Yet, the pupils’ use or lack of the MT was interpreted rather differently from the MT teacher’s perspective. Example 2 displays the contrast between the teacher’s interpretation of a pupil’s lack of contribution in Turkish during interactions and the pupils’ account of the same issue. Notice that the teacher generalizes from the experience that the pupil’s lack of contribution in Turkish is due to her negative attitude towards the MT. This example is taken from another field note where the fieldworker, Özgün Nergiz (XON), reports of the lack of Turkish competence among Turkish MTE pupils: Example (2): “What did you say?” Fieldnote by XON; 14-11-2013. Inden Mehmet kommer får jeg talt med Selda igen på gangen. Hun står med Gül (…) Selda fortæller mig da Gül ikke svarer på mine spørgsmål på tyrkisk, at hun ikke kan tyrkisk. Jeg taler med Mehmet om dette, som fortæller mig, at hun har fulgt tyrkisk en lang periode uden at lære så meget som et ord. Han giver et eksempel med at han siger ”Selda gel” ((Selda kom)) og når sætningen bliver gentaget for Gül, siger hun bare ”hvad siger du?” Mehmet mener klart, det skyldes hun ikke gider tyrkisk. [Before Mehmet comes I get to speak to Selda again in the corridor. She is standing with Gül (…) Selda tells me, since Gül does not answer to my question in Turkish, that she cannot speak Turkish. I talk to Mehmet about this. He tells me that she followed the Turkish class for a long time without learning a single word. He gives an example by saying “Selda gel” ((Selda come)) and when the same sentence is repeated for Gül, she only says “what did you say?” Mehmet believes that this is obviously due to the fact that she does not like Turkish.]

54

In this extract, the fieldworker points to two parallel interpretations about a former MT pupil’s (Gül) command of Turkish: first provided by Selda and secondly by the teacher. Selda, who is positioned as a Turkish speaker and communicate with the fieldworker in Turkish, comments that her friend (Gül) does not speak Turkish. In response to the fieldworker’s enquiry about Gül’s command of Turkish, the MT teacher provides his understanding. Mehmet demonstrates his evaluation in a short performance (Goffman 1956) for the fieldworker, who is a Turkish speaker himself. Mehmet first addresses Selda by a simple imperative sentence in Turkish (“Come here”), and repeats it, addressing Gül. In response, the girl asks him in Danish what he said (“What did you say?”). The two accounts by Selda and Mehmet are provided by in-group members of the MT class on a former class member’s (Gül’s) lack of command of the alleged MT. Whereas the first account is free from judgment, the teacher’s is far from that. In addition, he seeks to prove his point through demonstration and claiming that this is due to the girl’s dislike of Turkish: “she does not like Turkish”. Presumably, Mehmet bases his report on his teaching experience. An implication of this example is that the teacher is aware that in many cases, the pupils assigned to MTE are not necessarily speakers of Turkish – although it is assumed to be their MT. Sometimes they have to learn it at a very basic level, yet whether this endeavor succeeds or not depends on their interest. Accordingly, the Turkish teacher notes that Danish, for the Turkish MT pupils, is the dominant language and the language they prefer, as mentioned in an interview: “and among each other [the pupils’] first language is Danish (…) It is obvious to see (.) if one is satisfied with the Danish language”.xvi Also building on his perception (“it is obvious to see”) of Danish as the language the pupils prefer to use, he concludes that the pupils are content with Danish. At the same time the children’s satisfaction with Danish may imply the issue of sufficiency; that is, the Danish language fulfills the children’s communication needs. Thus from Mehmet’s point of view, the pupils do not desire (or need) a language other than Danish; they are not dependent on Turkish and show no affiliation with or interest in Turkish. Does this suggest that he questions the raison d’être of MT classes when the class agenda

55

insists on other patterns of language use than the ones the pupils already consistently use? Does it hint at his unrewarding insistence on using MT among the pupils? Whatever the fact is, the teacher admits that the patterns of language use by MT pupils are at odds with the presupposed and preferred patterns. Furthermore, while the MT teacher evaluates the pupils’ language shifts from Turkish to Danish, or their use of Danish as a sign of disinterest, he is aware of the fact that these pupils have Danish as their dominant language vs. Turkish as “mother tongue” in peer interactions as well as in their everyday language use. “A new language they have to learn” In the second set of examples, we see comparable patterns of language use and similar attitudes among the MT pupils and adults in the Arabic classrooms. The Arabic MT teacher’s observation and interpretation of the patterns of language use was rather nuanced when compared to the accounts reported from the Turkish class. The MT teacher in Arabic classes, Aslan, was originally a teacher of Danish for “bilingual pupils” (“tosprogslærer”) – that is, pupils who are regarded as having another language than Danish as their L1 or are considered as weak in Danish. As such, he had to assist pupils with their lessons and homework in ordinary (Danish medium) classes. Aslan’s attitude towards understanding MT pupils’ language use inside the Arabic classes came forward in a qualitative interview conducted by MSK. In response to a question about what the biggest challenges for a MT teacher were, he named three: 1) that Danish is the dominant language in the children’s everyday lives and even at their homes; 2) that some parents neither support MTE nor take it as seriously as the ordinary classes; and 3) that some children are not active speakers of Arabic and only the fact that the parents come from the Arabic countries of origin has motivated them to send the pupils to learn Arabic.xvii

56

Aslan’s remarks suggest that the children are exposed to different amounts of Arabic, and it is likely that this results in a class of children with highly different linguistic levels. Likewise, he evaluates Danish to be the dominant language in the children’s everyday lives. In his view Arabic, as a school subject, is not taken seriously by the families and/or they do not encourage the children to practice Arabic very often; departing from these points we expect to meet a diversity of patterns of language practice and use among the pupils inside and around the MT classes. In addition Aslan’s third comment about the challenges inside the MT classes pictures what is taken for granted, and the automatic relation of the pupils to Arabic as a consequence of inheritance and family background. This excludes in turn the pupils’ direct relation to both the heritage language and country and questions the institutional as well as traditional concept of nativeness (Leung et al. 1997) of Arabic regarding these children. Later Aslan nuances the reply with the following comment: Example (3): “The fastest way to communicate” Interview with Aslan by MSK; 24-06-2013. børnene er jo (.) selvfølgelig griber til den hurtigste vej at kommunikere med med hinanden og med deres søskende i hvert fald og det er jo dansk de:t fordi de kan forklare sig det går hurtigere det går nemmere (.) for dem (.) ø::hm (.) og så pludselig som det arabisk bli’r ik deres modersmål men det bli’r lidt æh æh ee et et nyt sprog de skal lære og (.) hh og det kræver li:dt mere (.) energi og arbejde lidt mere på at få dem til at (.) hvert fald få dem til at kommunikere (.) med hinanden (.) og med mig (.) på arabisk og (.) og mange af børnene sir også at de:t øh at de snakker arabisk med deres forældre men det med deres søskende er jo knap så meget arabisk. [children are (.) of course they take the fastest way to communicate with with each other and with their siblings at least and that is ((in)) Danish because they can express themselves it runs faster it runs more easily (.) for them (.) u::hm (.) and then suddenly Arabic becomes no more their mother tongue but it becomes a bit uh uh a a new

57

language they have to learn and (.) hh and it requires ((pedagogically; my comment)) a little more (.) energy and work little more pedagogical to make them (.) at least make them to communicate (.) with each other (.) and with me (.) in Arabic and (.) and many of the children also say that i:t uh that they speak Arabic with their parents but the language with their siblings is hardly that much Arabic.] In this extract Aslan explains that for the Arabic MT pupils, Danish is a faster mode of communication and thereby another form of pedagogy in MT classes is required. The pupils use Danish by default inside the class and with peers and siblings; this might be an influential parameter in the pupils’ rejection to use Arabic. Aslan’s account that “suddenly Arabic becomes no more their mother tongue” - presumably inside the MT class - indicates that Aslan describes an immediate change of expectation or assumption regarding the class pupils, as one might have considered Arabic the children’s MT. Accordingly, he views Arabic as a “new language” for these pupils. What Aslan reports of seems to be a shift from an ideal picture of MT pupils who are expected to be competent in Arabic to the picture of speakers with a higher proficiency in Danish, which is congruent with Aslan’s comments which I discussed before in example 3. This entails that there are expectations around the Arabic MT pupils and their relations to Danish and Arabic practice which are in accordance with the institutional labeling of MT classes. But the pupils’ patterns of language practice inside the MT classes, at their homes and in addressing different speakers demonstrate their affiliation to and proficiency in Danish vs. Arabic, regardless of their inheritance and family relation (Leung et al. 1997). In this sense, the pupils are proficient users (Paikeday 1985) of Danish rather than native speakers of Arabic. This is illustrated in the following example from a class interaction. The class practices an exercise where the pupils match a few flags with the names of the Arabic countries. Each pupil has to choose a country and find its flag:

58

Example (4): “Egypt” Class recording; Participants: MOH (Mohsen), DUH (Duha): pupils, and ALN (Aslan): MT teacher; 07-05-2013. Arabic, Danish, English. 01 MOH: 02 DUH: 03 MOH: 04 ALN:

05 UNI: 06 ALN: 07 08

[og Palæstina også taget] [and Palestine ((is)) also taken] [må jeg ikke få Egypten] [can I not have Egypt] det er også [taget] it is also [taken] [Fal]stin akhadhnah embareh [Pal]estine we took it yesterday Egypten Egypt masr Mohsen akhadha Mohsen has taken Egypt bil’arabi arjukum bil’arabi in Arabic please in Arabic masr akhadh Mohsen Mohsen has taken Egypt

In this excerpt, Mohsen and Duha take turns in addressing Aslan, both in Danish (l. 01-03). Aslan replies in simple sentences in Arabic, first without interrupting the interaction flow (l. 04 & 06), then he asks the pupils to shift to Arabic: “in Arabic please in Arabic” (l. 07). His instructional project in this sequence is to involve the pupils with a cultural activity about the Arabic countries, the related flags, and capital cities. But at the same time, he also focuses on language, both in terms of language choice and with respect to the linguistic constructions; he uses simple Arabic constructions, which are syntactically parallel to the pupils’ Danish question (l. 06 & 08). He topicalizes the name of the country in question “masr” (Egypt) and continues the interaction with an informative phrase of subject + verb form (l. 06) and then of verb + subject construction which is more in agreement with the Arabic syntax. Then he overtly encourages the pupils to speak Arabic (l. 07). By doing so, the interaction between the teacher and pupils is disturbed very little. The task is not only a cultural task - but a means to teach the Arabic language. Aslan does not ignore either the children’s fast intervention in Danish, but leads the interaction 59

while highlighting that the expected language is Arabic. Also he assists them in constructing simple sentences as if he is teaching Arabic to foreign language learners. There is another challenge connected to learning, using, and communicating in Arabic inside the classes. It concerns a discrepancy between the Arabic used in MT pupils’ families - called “dialects” - and the Arabic used inside the classrooms. During an interview, Aslan comments on the use of dialects vs. the standard language that many families speak: Arabic but they talk very much u:hm local dialect or rather a dialect which is not widespread, and the children have a very small vocabulary u:hm in everyday xviii language (Interview with Aslan by MSK; 24-06-2013) Aslan mentions the pupils’ lack of proficiency in using Arabic during their everyday lives in continuation of other linguistic aspects describing their families; the families’ use of local Arabic dialect and even “smaller” dialects which are not “widespread” and children’s small vocabulary. This is however not stated whether Aslan considers any causal relationship between dialects and the children’s small vocabulary. Also his remark that some dialects are not “widespread” is unclear. Does it mean the dialect speakers in general have less of a chance to communicate in their dialects within their network in Denmark? Or does he evaluate certain dialects as more isolated in general? Or does he refer to them as having more restricted vocabulary than other dialects or the standard language? Considering any of the three questions, which we saw in Section 3.2, Standard Arabic was not a MT of any of the pupils. In order to overcome the discrepancy between the Arabic taught and the dialects spoken in the homes, Aslan attempts to regulate language use inside his classroom. And to make children communicate in what he refers to as Arabic (which is preferably in Fusha) with him and their peers, as example 4 shows. I return to the Arabic varieties within home and school in Analysis II.

60

A few parents recognized and approved of different patterns of Arabic language use with reference to various situations of interaction. This is demonstrated by the following field note from an interview with two parents: Example (5): “Because it was the only language they could communicate together in” Field note by XLN on an interview with Hossain’s parents; 0110-2013. Jeg spørger også moren ind til hvordan børnene kommunikerer med jævnaldrende familiemedlemmer og hun fortæller mig, at Hossain snakker arabisk med den familie ((hans fætter; my comment)), som bor i Tyskland, fordi det er det eneste sprog ham og den jævnaldrende dreng kan kommunikere sammen på. Hun fortæller dog, at drengens arabisk ikke er særlig godt, da hans mor er halv tysker og først lærte arabisk senere hen i livet, hvorfor hendes søns arabisk er tilsvarende. Hun fortalte, at Hossain også havde grint lidt af ham engang imellem pga. hans udtale. Hun er selv glad for Hossains arabiske og at han ikke har en dansk dialekt eller lignende. Hun forklarer, at han i DK snakker arabisk til sine forældre, med et par danske ord engang imellem, men dansk mellem sine brødre og dansk med vennerne og fætter/grandfætter i Brønshøj. Dog taler han arabisk til morens veninde. [I also ask the mother how the children communicate with their peers inside the family and she tells me, that Hossain speaks Arabic with the family who live in Germany, since it is the only language he and the cousin of the same age as his can communicate in. She also says, that the boy’s Arabic is not particularly good, as his mother is half-German and only learned Arabic later in her life, due to which her sons’ Arabic is similar ((to hers)). She tells, Hossain also had laughed a bit at him once in a while because of his pronunciation. She herself is happy with Hossain’s Arabic and that he has no Danish accent or the like. She explains, that he in Denmark speaks Arabic

61

with his parents, but sometimes with some Danish words, but Danish with his brothers and with his friends and cousin and parents’ cousin ((my comment)) in Brønshøj. However he speaks Arabic with his mother’s friend] On the basis of what the mother reports, Hossain uses the two languages he has access to - Danish and Arabic - in accordance with his communicative needs and aims. He speaks Danish with peers and siblings in Denmark, Arabic with peers who do not know Danish, and Arabic with his parents and parents’ friends (presumably with Arabic as their dominant language). The mother also points to Hossain’s way of speaking Arabic regarding it as acceptable that Hossain inserts Danish features in his communication with the Arabic-speaking adults in Denmark: while he does not have a Danish accent in his Arabic, he sometimes switches to Danish codes. The mother distinguishes between two levels of mixing Arabic and Danish. While she views it as okay that Hossain inserts Danish lexical items inside his Arabic, she demonstrates to be satisfied with the fact that her son’s Arabic is not influenced by the Danish prosodic system, which is reflected as “Danish dialect” by the field note. This latter point most likely suggests Danish phonetic traits, which involve accent rather than dialect. Her concern is therefore in line with the adult participants’ concern, which I experienced in the Farsi classes, regarding their children’s Farsi being affected by a Danish accent. In this sense, the mother’s satisfaction that Hossain’s Arabic is not affected by Danish pronunciation maybe is due to her ideological preference of keeping the two languages phonetically separate. Her remark on Hossain’s mixing Danish with Arabic features indicates to some extent the mother’s acceptance of the phenomenon of polylanguaging (Jørgensen 2008; Jørgensen et al. 2011), which is language users’ usage of diverse linguistic features they have access to for communicative goals. Perhaps she views and interprets Hossain’s use of both Danish and Arabic lexicon during his interactions as a sign of privilege - i.e., having access to the two languages. However, any discernible Danish accent indicates that the MT has been influenced by the Danish phonetic system. A similar norm is reflected by her regarding Hossain’s attitude towards his family peer who does not have a good command of 62

Arabic or the accent in question. Hossain laughs at his halfGerman cousin’s Arabic pronunciation, which presumably is influenced by German. This illustrates Hossain’s awareness of particular social and linguistic norms inside his family and Arabic speakers, concerning the usage of a language related to a family’s origin. Following this norm, an affected way of speaking is devalued; the language of origin has to be kept “intact” and without the influence of “foreign” linguistic features. Fieldwork observations from different sites showed that parents’ understanding and expectation of the MTE pupils’ use of language in diverse interactional situations varied. In the following section, I present corresponding data from the project’s fieldwork in the Farsi sites. “Whether we want it or not - it’s our mother tongue” Inside the Farsi MT classes, different patterns of Farsi and Danish (and in some cases English) were used or preferred, which corresponded with the linguistic patterns studied in the Arabic and Turkish MT classes: The Farsi MT pupils primarily spoke Danish with their siblings and classmates. Their language of communication with the teacher(s) varied, depending on their command of Farsi, as the pupils had rather different levels of Farsi expertise. Many pupils almost always addressed their parents in Danish. A few parents replied in Danish or Farsi, while a few other reminded the children of formulating their comments or questions in Farsi instead of Danish when in the Farsi class or with other Farsi-speakers. In an interview, the teacher stated that he always reminded the families that they “MUST speak Farsi” at homexix, but not all parents did so (MSK’s interview with Mansour; 17-09-2013). Many parents and the main teacher explicitly asked pupils to reformulate their statements in Farsi, e.g. by saying, “Say it in Farsi” (lit., “be Farsi begu”). Yet, Mansour generally presented an indirect strategy for having the pupils switch from Danish (or English) to Farsi: either he corrected the pupils and prosodically emphasized the corrected form or he repeated their contribution in Farsi following a short silence. Example (6) displays the former strategy, where Laura (11 y.) utters a non-Farsi word in response to Mansour and becomes corrected by him:

63

Example (6): “Perfekt” Class recording; Participants: MAU (Mansour): MT teacher, and LAU & ROY (Laura & Roya): pupils; 19-04-2013. Farsi, Danish, English. 01 MAU: šomâ tamum šod (.) bacehâ you is it finished (.) children 02 LAU: [na] va (.) [no] and (.) 03 ROY [na] dârim mixunim dobâre [dârim mixunim] [no] we’re reading again [we’re reading] 04 LAU: [hey mixâym] barâye xodemun bexunim [we want to] read it for ourselves 05 tâ perfektDa-Fa bâše until it becomes perfektDa-Fa 06 (2) 07 MAU: bedune qalat bâše it becomes without mistakes 08 ROY: ˚âre:h˚ ˚yeah˚ The teacher-pupils’ interaction is in Farsi in this excerpt. Mansour (MAU), who has assigned a text to the two girls (Roya and Laura) to practice together, enquires whether they have fulfilled the task (l. 01). Subsequent to Roya’s response (l. 03), who informs they have not finished the task yet, Laura nuances the answer and justifies why the task has taken time, and that they want to practice more to master the text. She uses a non-Farsi word during her reply: “perfekt” (l. 05). Subsequently, Mansour takes a relatively long pause based on Laura’s construction, corrects her, and emphatically provides her with a respective phrase in Farsi “bedune qalat” (l. 07), which is received by Roya (l. 08). The point is that the adjective “perfect” is frequently used by Farsi speakers not only in everyday interactions but during their interactions on social media (my evaluation). Besides, the word has the same stress on the last syllable in Danish as well as in Farsi, with a phonetic difference – the /r/ is articulated in Farsi as a trill-alveolar /r/, which is the way Laura articulates it. Thus her use of this lexical feature follows the Farsi phonetic system even if the lexical item may be considered a loanword from Danish or even English. In fact, even if

64

we regard the lexical item in Laura’s reply a word originally borrowed from English - since it is broadly used in Farsi outside the Danish borders too - it will not change the scenario in Example (6), as loanwords in general become farsicized while being used by Farsi speakers. Example (6) shows how a non-preferred lexical item from another language is taken up by the teacher, and demonstrates the strategy he applies to substitute it with a Farsi formulation in the pupils’ utterance. The teacher’s remark functions as a metalinguistic comment and acts like a rupture within the interaction flow. However, his long pause can be interpreted in different ways. This may be due to the teacher’s strong concern about the children’s motivation to follow the MT classes; his attempt to ask the two girls to pay attention to his suggested word; or simply to find the best Farsi equivalent for the word. In any case this has generally been observed as one of the strategies to encourage the pupils’ use of Farsi without inserting foreign elements or accented lexical items. Pursuing pedagogical objectives or being motivated by pure language ideology, the teacher’s correction focuses on a monolingual language use model inside the class (see Analysis (II); and Karrebæk & Ghandchi (2015) for a discussion on ideological linguistic norms inside Farsi classes). Yet, during the sessions, when children addressed me (NG), who was the assistant teacher, their choice of language was Farsi or Farsi mixed with English, Danish, or exclusively Danish or English (if their language command was restricted). MSK noticed this variety of language level inside the Farsi classroom in a field note: There is obviously a large difference between ((the pupils’ and NG’s)) Farsi-repertoire. The little boy sounds as if he can say almost nothing, and NG translates all the time into Danish. His brother is not so very good either. On the other hand, MEH seems to have both much better receptive and productive competence.xx

65

Apart from this, parents in the Farsi sites had various attitudes towards their children’s patterns of language use in different periods of their lifespans. In some cases, their attitudes varied within a more detailed enquiry: when asked, they started with a general observation of the child’s language use, which varied in terms of situational and interactional needs, then reported on the child’s language use during different periods. In this sense, the description of children’s language use did not refer to a static phenomenon and was subject to change in different phases of their Farsi acquisition, and connected to the contact of Farsi and Danish when the children entered Danish institutions, the dominance of Danish in the children’s everyday life, as well as within interactional settings. Shortly, the parents generally provided both a synchronic and diachronic description of their children’s use of Farsi or Danish. However, they reported their preference of having children able to express themselves in Farsi as well as Danish. In an interview with a mother whose son, Shervin (7 y.), understood Farsi to some extent but used the language very rarely, the mother told how Shervin’s Farsi command decreased when he started using Danish in the institutions: Example (7): “Then it changed gradually” Interview with Shervin’s mother (MOT) by NG; 04-05-2013. Farsi, English. 01 NG: vaqtike (.) øøø hamun tâze when (.) uhh he was just 02 be donyâ umade bud cejuri bâhâš born how ((in what language)) did you 03 sohbat mikardin speak with him 04 MOT: mm fârsi sohbat [mikardim] mm [we spoke] Farsi 05 NG: [faqat] fârsi sohbat mikardin [only] Farsi you spoke ((with him)) 06 MOT: âre âre yea yea 07 NG: uhum (.) ba’d ârum ârum taqyir kard uhm (.) then gradually it changed 08 MOT: âre (.) hamun ke (.) mahdekudako inâ 66

09 10 NG: 11 12 MOT: 13 NG: 14 MOT: 15

yea (.) as soon as (.) he started šuru kardo= going to nursery and=az cand sâlegi rafte mahdekudak (.) how old was he when he began nursery daqiqan (.) precisely taqriban dah mâheš bud he was almost ten months uhum uhum širxârgâh raft (.) ba’dešam ke he went to the nursery (.) and then ba’daz unam (.) mahdekudako inâ after it (.) to kindergarten etc

The mother recounts Shervin’s language acquisition from the beginning (l. 04 & 06). The parents first communicated in Farsi with him until the age of ten months. Then, the communication model shifted when he started being immersed into Danish speaking environment. The institutional term of “MT learners” in Denmark is referred to as “bilingual” pupils who do not encounter Danish until they enter the Danish institution (UNI·C Statistics & Analysis for the Danish Ministry of Education 2008). At first glance, the mother’s description of Shervin’s language acquisition matches the definition of the Danish Ministry of Education. Yet the definition seems to be rather idealized, as other factors can be influential pertaining to children’s exposure to Danish in different families. In Shervin’s family, for example, his sister (18 y.) did not often use Farsi at home. Although the parents had spoken exclusively Farsi with Shervin since his first day (l. 04), Danish was already a linguistic means inside the family and when Shervin entered the nursery, Danish was not a new language for him. Elder siblings in such cases are a source of language socialization for the younger children in families. Besides, we cannot deny that children in migrant families - like in the case of this example - are exposed to Danish through Danish-speaking mass media. Moreover, MT children vary in their age of meeting Danish due to the families’ employment status. The mother in this example named her employment as an influential factor in Shervin’s attending the nursery. All these demonstrate certain assumptions around the 67

space-language relationships with regard to MT pupils, at both institutional and family levels. Places like home, nursery, kindergarten, and interactional spaces like in family (with parents, siblings, and relatives or acquaintances) or outside in the Danish society are taken for granted in relation to MT children’s exposure to and use of either of the languages. In general, like Shervin’s mother, a number of families in the Farsi sites recognized that a cultural and linguistic mix was a natural consequence of living in the Danish-speaking environment. They also noted that learning Farsi was required for maintaining transnational family contacts. This was also the case with Mehran. Mehran - regardless of a strong command of Danish and Farsi had used the languages to communicate with his paternal family who lived in Denmark and his maternal family residing in another European country. Among different members of his family, the older and adult members had a weaker command of Danish, which caused him to learn and use Farsi more actively. In an interview, when I asked his father (2nd generation migrant) who had made the decision for Mehran to attend the Farsi MT class, what the family’s reason was for that, he answered the following: Example (8): “Whether we want it or not it’s our mother tongue” Interview with Mehran’s father; 17-04-2013. râstešo bexâyd bištar mâmâneš (.) bištar mâmâneš vali xob øø vaqti bâ ham sohbat kardim (.) moteva- fahmidim kelâse fârsi hast barâ bacehâ hamin nazdikemunam hast (.) xob manam didam na xob dorost mige zabâne mâdarimune belaxare (.) bexâym naxâym zabâne mâdariye (.) avalan ke zabân (.) harci bištar adam (.) bald bâše behtare (…) barâ hamin bud (.) barâyeke zabnâne mâdarišo hoviy- rišeye asliyate irâni budanam (.) vaqti fekr kardim didim xob vaqti âdam zabâne mâdarišo balad nabâše (.) ye zare saxt miše ke be musike irâni be farhange irâni betune alâqe peydâ kone. [to tell you the truth his mom (.) mostly his mom ((encouraged him)) but well when we talked (.) I figured ((out)) understood that there was a language class for the children in our neighborhood (.) well I saw she was right 68

it’s after all our mother tongue (.) whether we want it or not it’s our mother tongue (.) firstly as for language (.) the more ((languages)) (.) you know (.) the better (…) that’s the reason (.) so that he could learn his mother tongue his identi- the roots of his Iranian origin (.) when we thought of it we saw well when one does not know his mother tongue (.) it would be a bit difficult to be able to develop a desire to understand Iranian music Iranian culture] In this excerpt, the father first provides a background narrative about what motivated him to send Mehran to MT classes. Having been encouraged by his wife (a second generation migrant too), he presents her argument about the meaningfulness of the MT, which he claims to have agreed with: “I saw she was right. It’s after all our mother tongue”. As mentioned earlier, Leung et al. (1997) have suggested three criteria to take into account regarding the relationship between language and language users, two of which correspond to the father’s narrative. He identifies himself (as presumably so does his wife) with Farsi and its associated culture. But he also pinpoints a particular direct correspondence between being a Farsi speaker and having national roots, and being originally Iranian to the identity of one with Iranian background. This account invokes the second criterion, language inheritance, according to which an automatic relationship ties individuals to a language as a consequence of birth and family background. The father’s argument in the excerpt is presumably provided in relation to both parents’ language attitude, but also stresses the two parents’ preferences considering their children’s learning of Farsi. Both parents have been raised outside of Iran. Nonetheless, the father’s account for the language as “our mother tongue” sounds as if he is connecting his family to a collective number of people with Iran as their background, and perhaps to the larger community of Farsi speakers in general. The father also nuances his argument by adding, “whether we want it or not” to “it’s after all our mother tongue”, which sounds like a firmly well-established relation determined by a taken for granted idea. This type of connection might have originated from (if not been imposed by) different sources, such as the viewpoint of the in-group Iranians, regardless of their birthplace and length of residency in a second 69

country. This can also be imposed by the host society with certain language ideologies around linguistic minority groups - this is however beyond this study and requires a further research. Whatever the source of this ideological relationship is, to this parent the similar connection is a key to know the culture of one’s country of origin, ergo one’s identity and heritage roots. Finally, the “desire” to understand the ancestral culture (e.g., Iranian music) is developed through the Farsi language in the father’s view. This provides an interesting example of how an individual’s MT as a presupposed term can more accurately fit into the definition of “heritage language,” that involves an ancestral link across migrant generations and a desire for heritage cultures and roots (2.2.2.). The pupils’ relationship to Farsi, being Iranian, and having Iranian families often gave them a tinge of their families’ attitudes whenever they meta-discursively pointed to those issues, as we see in the following.xxi Example (9) is a sequence from the Farsi class in which three pupils, Parsa (7), Mehran (10), and Amir (5), are actively negotiating such issues in Danish. Parsa has a very restricted command of Farsi and understands only a few single Farsi words. Amir communicates in Farsi at home, according to his parents; however, he resists using Farsi in the classroom. According to his brother, Mehran, Amir is fluent in Farsi (l. 09). The sequence in Danish is initiated by Parsa asking NG a question about Amir’s Farsi speaking: Example (9): “I’m also Danish” Class recording; Participants: PAR (Parsa), AMI (Amir) & MEH (Mehran): pupils, and NG: fieldworker; 30-11-2013. 01 PAR: er han dansker eller farsier (.) is he Danish or Farsish (.) 02 eller er han begge dele or is he both of them 03 NG: spørg ham selv ask him yourself 04 PAR: kan du farsi eller begge dele (1) can you speak Farsi or both sides (1) 05 eller dansker (.) eller begge dele or Danish (.) or both sides 06 AMI: danskDani07 PAR: men du kan ↑tale (.) persisk (.) 70

08 AMI: 09 PAR: 10 AMI: 11 MEH: 12 NG: 13 AMI: 14 PAR: 15 16

but you can ↑speak (.) Persian (.) jah (1) haha yeah (1) haha kan du tale flydende can you speak fluently ((he shakes head)) [↑jo du kan] [↑of course you can] [er det derfor] du er dansker [is it why] you are Danish ((he nods)) jeg er også dansker jeg vil (.) I am also Danish (.) I want (.) min far sagde bare jeg skulle my dad just said I had to lære persisk … learn Persian …

This example illustrates that Parsa is curious to know what type of language user Amir is, probably after observing that Amir does not participate in the new exercise. Parsa does not build his enquiry on Amir’s language command, but rather on a simple and stereotypical binary of being Dane (i.e., “dansker”), which would exclude being Iranian (i.e., “farsier” in his word). Instead, he provides a third and hybrid category, “both of them” (l. 02). Following my invitation to ask his question directly to Amir, Parsa reformulates his question on being Dane, Iranian, or both through being speakers of either of the three types (l. 04-05). This is received by Amir this time and answered shortly in Danish (l. 06). Parsa does not agree with this response, though, so he insistently disagrees and objects to Amir’s utterance to him, saying that Amir “can ↑speak Persian”. This objection brings about Amir’s reaction, as he playfully expresses his affirmation that he can. But in response to a more nuanced attempt by Parsa, Amir says that he does not fluently speak Farsi, which is rather different from Mehran’s viewpoint. Being interested in making the responses more precise, I ask Amir whether his not being able to fluently speak Farsi is the reason for his being a Dane. He nods, and accordingly, Parsa takes the same turn and gesture. Parsa also adds that, for him, sitting in the Farsi class is his father’s decision; his father makes him learn the language (l. 15-16). xxii What is

71

noteworthy in this excerpt is that, firstly, these pupils relate speaking a MT or another language (i.e., Farsi and Danish) to a group of people who are bound by conventions to a country label, Iranian or Danish (in his words “farsier” and “dansker”). Secondly, one can be regarded as a legitimate speaker of a particular language regardless of whether she or he is fluent in that language. Thirdly, MT pupils can have a third possibility in their linguistic self-identification, which goes beyond a simple dichotomy of linguistic competence in Farsi or Danish, involving a hybrid status. And finally, being a pupil of MTE classes can only be a consequence of the MT parents’ interest or ambition, and does not necessarily have to do with the MT pupils’ linguistic competence per se. In sum, the examples from the three MTE sites illustrated different types of patterns of language use preferred by various participants. The MT teachers applied different strategies consequently to regulate the use of the MTs among the pupils and in the pupils’ reference to the teachers: claiming lack of command to understand Danish in answering pupils’ Danish interaction; a short pause subsequent to the pupils’ mixed language, coupled with insistence on or reformulation of the related words in the MT; and a constant interaction in the two languages. Yet, the teachers similarly asked the pupils to speak in the MTs; they noted that other languages like Danish or English were not preferred inside the classes and that children had to formulate their comments, questions, and answers in their MTs. Thus, the MTE classes were reserved stalls (Goffman 2010) in the sense they were defined by concrete places, materials, and particular language speakers (participants) with shared socio-cultural backgrounds. In other words, the classes were not only the concrete (and sometimes unstable) rooms surrounded by walls mostly covered with irrelevant posters in other languages (see 4.2.), and by a door that created a concrete space specified for teaching the MTs – they were also associated with norms of language use which were predefined and explicitly or implicitly reminded to the MT pupils. The use of languages - MTs, English, and Danish - in these classes appeared as being defined in relation to spatial boundaries. The three teachers of Turkish, Farsi, and Arabic MTs applied to the

72

norm of double monolingualism (Jørgensen 2012) with different classroom consequences. The pupils in the Farsi classes had to use one and only one language separately during their interactions with the teacher. This norm can be interpreted in Goffman’s terminology as a criterion with which the possessional territory of certain speakers can be identified. A language named MT, having certain linguistic features and properties and allegedly belonging to particular speakers, was claimed as the speakers’ possessional and defendable territory. Linguistic features associated with other languages than MTs entering this territory appeared as encroachment, and this was criticized by the MT teachers. Yet, the MT pupils had and continued to use other linguistic resources inside the MT classes, in their homes, and with their family members outside the borders of Denmark. I believe MT pupils treated the linguistic features of various “languages” differently. Motivations for such treatment of linguistic features by MT pupils are beyond the scope of this study, though. In what follows, I return to these terms for a further discussion in relation to MT pupils’ understanding of relating a “language” named MT (or Danish, etc.) to their and their parents’ countries of origin. 5.1.2. What language in whose country? In this section, I focus on a number of examples to illustrate how language users in the study partition places through each other’s language use, and how this partitioning is connected to sociogeographical territories. I also show how the geographical borders - or frontiers - change or extend through the various participants’ preferences of language use. Thus, I treat questions like the following: How does a language connect with geographical places? What language(s) do MT users (have to) speak in the countries of origin? What language(s) do they (have to) speak in Denmark? What countries are involved in the MT classes? I start this analysis with the Turkish site. “We spoke Danish in Turkey” In a field note written by JNJ, he reports of the linkage of languagecountry, within a dialogue between the teacher and the MT pupils about the pupils’ vacations. Prior to this sequence, Mehmet had 73

asked both Fadime and Özlem a number of questions, to which mostly Fadime had responded in short and incomplete Turkish sentences. By Özlem’s first turn, Mehmet addressed them as follows: Example (10): “We spoke Danish in Turkey” Field note by JNJ; 08-25-2011. Mehmet spørger, hvilket sprog de har brugt i Tyrkiet, og Özlem svarer på dansk: ”Vi har snakket dansk i Tyrkiet”. [Mehmet asks, which language they have used in Turkey, and Özlem answers in Danish: “We spoke Danish in Turkey”.] The above excerpt concerns two turns taken by Mehmet (in indirect speech) and Özlem (in direct speech). Throughout his field note, JNJ has chosen to voice the teacher through the indirect speech and the pupils through the direct speech. In a warm up exercise, Mehmet asks the two pupils in Turkish what language they spoke during their vacation in Turkey, to which Özlem directly and briefly replies in Danish that they spoke Danish in Turkey. A few motivations are possible regarding both turns. Mehmet might have assumed to receive an answer such as Turkish, seeing that this is the official and national language of Turkey. Or he might have assumed that the pupils have travelled to Turkey to visit families, and/or because that is the target language of the class and is expected to be used somewhere in the students’ personal lives. Another motivation might have been to start the new semester of the MT class talking about related linguistic experiences from the pupils’ country of origin. However, the response made by Özlem is at odds with both the previous short and incomplete answers in Turkish and with the preferred answer, and thus has an interpersonal impact. A number of questions can be posed in this regard. Does Özlem’s answer challenge an assumption of using Turkish by the alleged Turkish speakers within Turkish borders? And, if so, who would be making such an assumption? Or is it indexing the Turkish MT pupils’ probable affiliation to the Turkish-speaking group of people, which have Turkish language and culture in common? Özlem’s response to Mehmet appears as a challenge to certain 74

presuppositions. In a conversation with Özlem (19-12-2013), I asked her why she attended the Turkish class. She mentioned the importance of learning Turkish in two respects: First, Turkish was to her a means of interaction with her father due to his lack of Danish command. Second, Turkish was necessary for communication during their travels to Turkey. However, she thought that she would rather stay in Denmark, as she was used to being in Denmark where she was born, grew up, and went to school. This indicates that Özlem finds a need for knowing Turkish as a communicative means, both in her family and as soon as she finds herself in Turkey. In addition, recalling Example (10), she replies with “we” (i.e., the deictic plural pronoun) rather than “I” and this renders her as a member of a group of language users who used Danish in Turkey. Although one possibility is that she voices the concerns of her family, another possibility is that she answers on behalf of the pupils (e.g., Fadime) whom Mehmet addressed. In either case, being deictic in nature, the referential “we” taken up by Özlem certainly contradicts the group referential index that the MT teacher attempted to construct during his instruction; that is, Özlem’s reference certainly seems at odds with the usage of “we” by Mehmet in various settings, such as in the following data taken from a field note written by fieldworker Nergiz (XON): Example (11): “We start ((the week)) with Sunday” Fieldnote by XON; 05-12-2013. Mehmet ender som altid med en forklaring af at i den europæiske tradition starter ugen rigtig nok med mandag, men hos ”os,” jf. tyrkere, starter vi ((ugen)) med søndag. [Mehmet ends as usual with a explanation that in the European tradition the week starts let’s say with Monday, but in “our” country, cf. Turks’, we start ((the week)) with Sunday] This excerpt illustrates that Mehmet compares two systems of weekdays, based upon the official calendars of Turkey and of other European countries. But throughout his accounts, by positioning the pupils as part of a community referred to as “us”, the teacher 75

aligns the MT pupils with a group of Turkish speakers. In fact, the reference “us” is interpreted as “Turks” by the Danish-Turkish fieldworker. Mehmet also aligns the pupils with Turkey as the geographical territory where another calendar is used, and by doing so, he contrasts a different system of naming weekdays that covers the rest of Europe. This type of socio-cultural and geographical alignment of the MT pupils with the country of origin (regardless of their place of birth – mostly Denmark) may be interpreted as an attempt to create an imagined social space. The social space involves inside-Turkey people as well as people with Turkey as their origin outside Turkey. In addition, it may be interpreted as an assumed expansion of the Turkish geographical borders into the Turkish MT class. In this way, the MT class is a more mundane type of embassy - in the same vein as a representative space of the related country - and regarded as a part of Turkey’s territory, within the walls of which linguistic norms, seen as characteristic of the Turkish territory, are expected by the teacher. Turkey, in this sense, is to some extent used in the Turkish class as a normative center (Agha 2007) towards which the MT teacher attempts to orient his class and based on which he organizes his instruction. I return to the issue of certain linguistic norms in Analysis (II). “We should speak Farsi” Many MT pupils often demonstrated a dis-preference to follow or stay inside the MTE classes, but not all MT pupils expressed negative feelings and dislike in relation to their country of origin. In one case I observed a rich point (Agar 1995) about the pupil Shadi (9 y.), whose mother reported of her dislike towards Iran. The interview with Shadi’s (SAD’s) mother took place after their travel to Tehran during the Christmas vacations. In response to my question on Shadi’s everyday habits of language use, the mother said that Shadi was a former fluent Farsi speaker (until 6 y.). The mother also explained Shadi’s negative attitude towards Farsi speaking, especially when at home. I asked her whether she thought Shadi was not interested in using Farsi or if she felt it was easier to use Danish rather than Farsi. This brought up the following sequence:

76

Example (12): “We have to speak in Danish” Interview with SAD’s mother (MOT) by NG; 10-02-2013. Farsi, Danish, English. 01 MOT: mas- ba’zi vaqtâ dust dare (.) for ex- sometimes she likes (.) 02 mige pinligdk-ye (.) ba’zivaqtâ says it’s embarrassing (.) sometimes 03 yebâr bargašt (.) yâdame do sâl piš once she turned to me (.) I remember 04 be man gofteš ke (.) she said about two years ago that (.) 05 ci(.) mâ- mano šoharam dâštim that (.)we- with my husband 06 Fârsi sohbat mikardim I was speaking Farsi 07 bargašt goft (.) hh then she turned and said (.)hh 08 vi bor i Danmark (.)hh we live in Denmark (.) 09 vi skal tale dansk we have to speak Danish ((She stylizes Shadi)) 10 NG: nå ja haha okay I see haha okay 11 MOT: haha man ye-zare goftam (.) haha I said a bit (.) 12 na sâket bâš (.) yani ce harfiye no keep quiet (.) what do you mean 13 mâ irâni’yim we’re Iranians 14 bâyad fârsi sohbat konim (.) we should speak Farsi (.) 15 to’am bâyad fârsi yâd- sohbat koni and you you have to lear- speak Farsi too This excerpt demonstrates that, to the mother, Shadi’s attitude towards Farsi is mixed. The mother also reports that sometimes it is embarrassing to Shadi to speak Farsi (l. 1-2). I think this feeling is situationally dependent, as the mother points to her contradictory attitudes towards it. The attribute “embarrassing” motivates a short narrative, focusing on Shadi’s reaction, when she once opposed her parents’ use of Farsi. The mother then performs the scene while voicing both herself and Shadi. The mother’s 77

performance implies that she criticizes Shadi’s attitude while asking her parents to stop the use of Farsi and to shift to Danish, because they are living in Denmark. She stylizes Shadi and accompanies her own account with laughter. The sequence dates back one year prior to the time Shadi started the Farsi class. The quote reflects Shadi’s awareness of certain social dynamics that tie communicative and other social spaces as well as geographical places to a language. It also reflects how she has announced her alignment with the idea of speaking Danish within the Danish borders. She protests her parents’ breaching certain assumed norms and violating an imagined linguistic boundary that might have been conventionalized socially. In fact, she asks her parents to remember this alignment and apply these norms. The mother reacts to Shadi’s comment by demonstrating her parental autonomy. She does so to indicate language users’ agentive role in choosing the language they identify themselves with, and perhaps to fulfill her generational mission in maintaining the language of the country of origin. Departing from this reaction, the mother reminds Shadi that she in turn is meant to learn and speak Farsi, as the family is Iranian. Thereby, the mother rejects any alignment asked by Shadi and points to the heritage and family links to Iran, instead. Apart from this, the challenge within the narrative in the above excerpt can also be read in the light of territoriality’s key elements - classification, communication, and enforcing control (Sack 1986). Two territories are linked together: one is what I call the interactional territory of a family conversation between the parents, in which Shadi takes her turn. This territory is bound to an overall social and linguistic space, which is linked to the territory of Denmark with its concrete borders and associations. According to the mother’s narrative, Shadi takes the relationship between to live in Denmark and to speak Danish for granted – in the larger society as well as in a migrant family linked to any MT. Denmark is introduced as a territory to which “we” (Shadi’s word) are linked. The linguistic resources associated with this territory and with other territories (e.g., Iran) distinguish between the related social groups. Thus, if “we” are located in Denmark, “our” language should be Danish and other languages should not be allowed. This

78

norm, which Shadi announces in Danish, restricts the family interaction. Shadi’s protest to the parents’ Farsi speaking represents displaying and enforcing control during the family interaction. Though, it is most probable that her being a teenager may not lend success in enforcing control over this interactional situation. And from the mother’s point of view, there is no such linkage between the larger territory of Denmark and the interactional space of her conversation with her husband. The intent of Shadi’s territoriality thus gets reversed as the mother takes advantage of her parental authority and announces her own territoriality over family interactions in Farsi as an immediate consequence of being Iranians. As Sack (1986) puts it, situational interactions within territories are impacted and administered by power, which is not necessarily explicit. Interestingly, in the excerpt above, the power relation exists within the territory and influences the territoriality claimed by the two interactants, but power itself is enforced by other social parameters such as generational contentions and identity construction. Pupils’ attitudes towards Denmark vis-à-vis countries of origin are mixed, as we see in the next section. In the continuation of this analysis, I turn to the Arabic MT classes. “Where is my country?” In an exercise in the Arabic classes, matching Arabic country names to the related flags encouraged some discussions about the relationship between language users and countries of origin. Example (13) represents the negotiation of possessive relations to the country of origin inside a short sequence from this class exercise. The pupils have to match the country names, written in Arabic on pieces of paper, with flags and capital cities. The sequence starts with Aslan addressing Mohsen and Dina while handing them their papers:

79

Example (13): “My country” Class recording; Participants: ALN (Aslan): teacher, DUH (Duha), INA (Iman), and MOH (Mohsen): pupils; 17-05-2013. Arabic, Danish, English. 01 ALN: Mohsen lailak la Dina Mohsen for you and for Dina ((Aslan gives the pupils each a piece of paper)) 02 DUH: hvor er mit land where is my country ((she means the country she was assigned to the previous session)) 03 INA: har du skrevet på [den] have you written on [it] 04 ALN: [↑mit land] [↑my country] ((he looks up and stares at her)) 05 MOH: det er ikke dit land it is not your country 06 ALN: det er ikke dit land it is not your country ((without looking at Duha)) 07 DUH: neej (.) men altså jeg kommer derfra no: (.) but well I come from there 08 UNI: hvor kommer du fra where do you come from Duha has not found the paper she had presumably taken the previous week, and thus asks Aslan where her country is (l. 02). This question is simple and innocent at first glance, but it has another meaning within a sequence that contains the turns taken by Iman (l. 03), Aslan (l. 04 & 06), and Mohsen (l. 05). Iman responds to Duha’s question as natural and perhaps expected, and asks Duha whether she has written the name of her country on the paper. Aslan’s repetition of and emphasis on the issue that Duha asks (i.e., “↑my country”) reflects his surprise of Duha’s understanding of her connection to the country in issue, while adding to its denotation or its reference. He additionally expresses this by a gesture of direct and questioning gaze. This is followed by Mohsen, who points out that the country in question is not Duha’s. Right after that, the statement is repeated by Aslan (l. 06) as a remark aiming presumably at correcting Duha’s remark (and

80

perhaps understanding). But in response, she insists on her position, and makes a clarifying comment that adds nuance to the interaction. Her comment consists of two parts. First, she affirms the others’ suggestion that the country is not her country by a stressed negation form “neej”; at the same time, she introduces a contrast by “but” (l. 07). In fact, the matter that she “comes from” the country (which is Lebanon) is represented as being sufficient for her naming the country as a possessional territory she is identified with socially or through her family. Whether Duha identifies herself with the country in question by inheritance or language ties (i.e., through her family or Arabic), she does not receive the position given by both Aslan and Mohsen. They emphasize the mismatch of Duha’s claim over a link with the country of origin and where she belongs to, possibly Denmark. However, “her” country (Lebanon) might be considered a part of her and many other Lebanese individuals’ possessional belonging when it comes to their identity work, as if she is saying she is anyhow related to the country of her parents. Besides, being born and raised in Denmark (which is her case) does not seem to contrast, from this view, to her relation to Lebanon. So, she justifies the position she has taken on for herself, regardless of the counter-claimants’ (in Goffman’s words; here, Aslan and Mohsen) endeavor to qualify her claim. To sum up, I move from the diversity of attitudes towards countries of origin and Denmark, as well as from the sense making out of links established between these territories and the related languages, towards the use of MTs and in-/out-group focuses on community construction through MTs. Through the examples in this part on language use and choice, we saw that certain territories at the macro level were reserved for MTs and were wellestablished, both from the institutions’ and in-groups’ perspectives. From the two perspectives, the pupils were ideologically treated as idealized “bilingual” and “MT” users, by virtue of having heritage links to some geographical territories. In many cases, the MT teachers drew their similar arguments from the countries of origin as centers of normativity to set their linguistic norms (I return to this in Analysis (II) to discuss what type of MT was preferred inside the MT classes). In addition, they

81

positioned the pupils from an in-group perspective, and acted towards engaging the pupils as members of MT-speaking groups. One possible explanation for such a perspective may be that the instruction acted as a means to create a social space for the construction of group identity through MTs and for the sake of maintaining that. But within the same well-established territories, the pupils demonstrated deviations from the presupposed norms inside the MT classes. They also introduced new territories of the situated interactions, within which they actively negotiated their own claim(s). They used mixed language within or around the MT classes and in their interactions with the adults; they attempted to link language with a related country through a different configuration, which was neither necessarily in agreement with the MT adults’ understandings and beliefs, nor with the institutional stereotypes regarding a given idealized bilingual and MT user or learner. However, particular cases illustrated that MT pupils can be or are affected by ideological attitudes that already exist in their linguistic and social environments. 5.2. Analysis (II) Mother tongue as a social phenomenon: The sociolinguistics of the MT classrooms This part draws on the third part of the thesis’s theoretical framework as it aims to study registers (Agha 2007) inside the MTE classes under study. With this analysis I demonstrate from an inside vantage point how certain preferred ways of speaking add to the meaning of an institution called mother tongue class (MT class). I focus on two major registers - standard and dialect of the respective MTs - and the associated features of the registers that researchers (including me) observed or were told about mainly by adult participants. The data is drawn particularly from the interviews with the adult participants and the classroom recordings. I explore three things:

82

(1) which of the two registers were present, taught, or (dis-)preferred in the MT classrooms; (2) whether the language users connected these registers to any territories; and (3) whether the taught or to-be-learned register was (counter-)valorized (ibid.), that is, whether participants labeled them as normal/abnormal, right/wrong, strange, and so forth.

5.2.1. “Normal Farsi” In this section I treat understandings of and evaluative attitudes to a “language” named Farsi that was taught to MT pupils. Given that the pupils’ families originally were from different geographical places and had various linguistic backgrounds from Iran, I intend to see what language register the adults preferred the MT pupils to learn to use and communicate in. I do so mainly on the basis of an interview with a mother to explore social stereotypes about Farsi, which are linked to Iran, but also based on my observations from in- and outside the MT classrooms. A year prior to the project launch, I noticed a pupil (6 y.) with a restricted command of Farsi who spoke in Farsi with a regional accent and in general preferred to answer in Danish. Her accent was associated with central Iran, where her parents came from. By the project launch, she left the class, and her father reported to me that despite his and his wife’s interest in sending their daughter to “mother tongue classes,” she was no more “interested in participating in Farsi class” (lit., “vey alâqemand be edâmeye šerkat dar kelâse fârsi nemibâšad”). When I inquired the reason through another mother (who knew the former family), among different reasons she named the accent of the child as a challenge. Of course, I cannot be certain about this as more than one reason normally exists for a pupil to quit a class. But this example made me aware of the way(s) of speaking in the MT class. The Farsi taught and spoken in the class was in the Tehrani dialect. However, this was not the most appropriate choice due to another challenge: spoken Farsi (almost regardless of dialect) disagrees with its written form provided in pedagogical materials. In the case of the pupils with less or almost no command of Farsi, the MT teacher focused on the written variety 83

to achieve the goal of the MT class, that is, literacy, which he mentioned on various occasions. But in the case of uncontrolled interactions (e.g., everyday conversations) the language shifted to the dialect of Tehran. This Farsi register has been regarded as the prestigious and standard form of Farsi and has functioned as a baseline for many Iranians (Perry 1985). During my interviews with families, I inquired about the families’ use of Farsi varieties, their attitudes towards the varieties as well as where they came from in Iran. Distinguishing among different varieties, some showed concern about their children’s choice of geographically localized varieties. They also reported whether and to what extent the pupils’ use of varieties was important to them. Example (14) displays this issue. It is taken from an interview with a mother who was born in southern Iran. Prior to this excerpt, she reported that several of her family members spoke with either an accent or dialect - she used one term (i.e., lahje) for both. Notice how she distances herself from the accent and dialect she knew from her family and wishes the same for her children. To answer my inquiry of whether the children picked up the regional accent or dialect when they visited Iran, the following response came up. Example (14): “I like they speak Farsi” Interview with Nasrin’s mother (MOT) by NG; 04-10-2013: Farsi, English. 01 MOT: haha bâ lahjehâye moxtalef cerâ (.) haha in different accents yes (.) 02 sohbat mikonan they speak ((different accents)) 03 vali xob man say mikonam dobâre (.) but well I try again to (.) 04 dobâre bâ xodam sohbat mikonan when they talk to me again 05 dobâre dorost mišan they become correct again (…) 18 NG: dorost mišan yani ((when you say)) become correct means 19 mesle šomâ sohbat mikonin like the way you speak 20 MOT: âre mese ↑man hmm mayeah like ↑me hmm ordi84

21 22 NG: 23 MOT: 24 (…) 49 NG: 50 51 MOT: 52 NG: 53 MOT: 54 NG: 55 MOT: 56 57 58 NG: 59 MOT: 60 61

fârsi mamuli be qole as we say the ordinary Farsi aha fârsiye haminjurixxiii (.) the common Farsi (.) na bâ lahjeye masalan not with the accent of ((such an such place)) âre ce ta- (.) ce eškâli dâre hâlâ yea what (.) what is the matter now bacehâ bâ lahje sohbat konan if the kids speak in another accent nemidunam I don’t know uhum dust dâram Fârsi sohbat konan (.) âre I like they speak Farsi (.) yeah âhâ unvax unvax (.) xodam besides besides (.)personally lahjeye A. ro hicvax I never liked dust nadâštam haha the A. accent haha âhâ okay (.) lahjeye= âhâ okay (.) the accent of =pedaramam hicvaqt A. sohbat nemikard my father never spoke with A. accent pedaram xeyli hhh xe:yli my father spoke in a very hhh ve:ry ketâbi sohbat mikard (1.) xeyli literary manner (1) very much ((the way one writes))

The extract illustrates that the mother is aware of the influence of (two) Farsi regional varieties related to her family in Iran on her children’s way of speaking subsequent to their family contacts, “they talk in yeah (.) in different accents” (l. 01-02). In fact she laughs in remembering it (l. 01). But she pauses, “but well” (l. 03) and contrasts the way of speaking her children picked up with another way of speaking that she relates to the time prior to the family visits and reports further how the children turn to the prior way of speaking as soon as they speak with her, “when they talk to 85

me again (…) they become correct again” (l. 04-05). She describes the two ways of speaking connected to before and after her children were influenced by the regional registers by the attribute “correct” (l. 05). In her further description of the “correct” way of speaking, she adds “ordinary Farsi”, “common Farsi”, “not with an accent”, and “like ↑me” (l. 20-23). I have chosen “common Farsi” (l. 23) as a translation of a Farsi adjective to use in the colloquial language which means “of such” or “in this manner” for the sake of fluency in translation. The adjective adds to the meaning of the first personal pronoun “me” (l. 20). Having deictic referents (Levinson 1983, following Peirce’s categorization of linguistic signs), both features anchor the preferred variety that the mother identifies herself with to the here-and-now context of the interview, presumably perceived by NG, as interviewer. By orienting to her own variety within the spatio-temporal frame of the interview, she indexes her own way of speaking as preferred which to me, being raised with Tehran’s so-called accent, is the variety of Tehran - and she wants her children to acquire and use it. Moreover this variety is opposed to the regional way of speaking she knows from her family (which comes up in two other parts of the interview), except for her father, who chose not to speak in the regional variety but instead in a literary way (l. 58-61). Regional varieties are grounded in geographical areas and are understood in contrast to the standard variety (Mæhlum 2010). In addition language users’ attitudes towards various registers are impacted by the history behind “languages,” and based on those attitudes, they appraise certain registers and reject certain others (Agha 2007). The mother in the example expresses her dislike about the two regional varieties associated with her family; she wants her children to pick up her own, which seems as the standard variety of Farsi; and valorizes the latter and former varieties by positive and negative attributes. She provides linguistic accounts for and describes her attitudes towards different users (including her children) of dialects versus a standard variety of the language. In this sense, her evaluation of the dialect and the standard register is metapragmatic (Agha 2007). Her naming of her father as an ideal example of using the formal register of Farsi (“literary” ketâbi in her words) may indicate as well that she has chosen

86

consciously the similar way of speaking and wishes the same for her children. Yet what matters here is first the differentiation and valorization represented in her account on varieties vis-à-vis the so-called standard form and, second, how she relates the latter variety to the formal social class with a higher status of access to the written language. In brief, the mother understands a dialect and an accent as varieties that are not “formal”, “fluent”, “official”, or “written”. In another part of the interview, she associates dialectal and accentual varieties with women and older people, and she does not wish her children to learn or use them. Although not mentioned explicitly, the issue that regional varieties are not codified renders them inferior to the literate social class. She also believes accent or dialect is “funny,” and thinking of such linguistic features makes her on several occasions laugh during the interview. In sum, building upon her contrast between the standard form and other varieties of the language, she constructs a linguistic dichotomy that indexically represents which part of the country she comes from and which parts she might wish to be identified with. Whereas I observed the dominance of Farsi based on the variety of Tehran and thus the standard variety, on some occasions I noticed that the main teacher, a Kurdish speaker of Iranian, communicated in Kurdish with a few families who stemmed from Kurdish regions in Iran (i.e., West Iran). However, this never happened in the classroom or in the margins of the classroom (i.e., in the corridor, after the class program, or during the breaks). Interestingly, although the Turkish teacher was also a Kurdish speaker from Turkey, he demonstrated a rather different attitude. In the next section I will describe a different situation we observed in the Turkish MT classes. 5.2.2. Turkish varieties in Turkish MT class In this section I analyze a few illustrations of attitudes in the Turkish MT class towards the standard vis-à-vis other varieties of Turkish with particular focus on a specific phonetic feature, named as “soft g”, written as ğ, and known in everyday language as yumuşak g. In my second visit at the Turkish site, I noticed that two pupils mocked the MT teacher. This occurred right after he had pronounced the soft g in several Turkish words. The two pupils’ 87

reactions made me aware of a tension inside the classroom between the teacher’s pronunciation and the pronunciation that seemed to be preferred by the children. The following vignette reflects that: Example (15): The pronunciation of soft g Field note written by NG; 20-02-2014. (Original in English) First Jaya and then Hadi make fun of Mehmet’s pronunciation of soft g each time he pronounces it. In fact he pronounces the words including soft g very similarly to Azeri-speaking Iranians, like /ɣ /. But I do not think he does so in purpose and on the contrary, it seems he does not care about pronouncing /ğ/. The field note highlights the teacher’s soft g as a significant feature in his instruction. It also demonstrates that the pupils are aware of the teacher’s way of speaking and this shows it represents a difference to the pupils’ own use, their families’ use, or perhaps to the received normative expectations. Yet the teacher, to my observation, does not do anything to change his way of speaking or account for the use of this pronunciation. In other words, the impact of the use of this feature in the class seems to be either not noticed or ignored by the MT teacher. The pronunciation of soft g depends on the letter’s vocal surrounding and the language users’ regional variety (Ünal-Logacev et al.: 2014). The pupils in the Turkish class are originally from different regions of Turkey and speak different varieties and thus treat differently the words characterized by soft g. Studies report that the use of soft g is considered significantly less-valued by a large group of Turkish speaking inhabitants of Western Turkey and have historical associations with a lower-status social and regional group of speakers (Demirci & Kleiner 1999). Inside the Turkish class in question, the phonetic trait creates an ideological tension and renders the teacher’s language a variety not recognized by pupils. The difference of the varieties of Turkish in the class, namely the one taught and spoken by the teacher and the other varieties that might be expected by the pupils, was brought up within another sequence in an interview with Hadi’s mother. Hadi is also present

88

and participates. Prior to this sequence the researcher Karrebæk has inquired whether the mother has ever met or contacted the teacher. In the following sequence the mother is asked what she thinks about the Turkish class. The mother has spoken to him only by phone: Example (16): “It was a bit strange” Interview with Hadi’s mother (MOT) by MSK & XON: fieldworkers; Hadi (HDI): MT-pupil, present too; 01-04-2014. Danish, English, Turkish. 01 MOT: (…) haha jeg synes det var lidt underligt … haha I think it was a bit strange 02 på den måde Mehmet snakkede the way Mehmet spoke 03 ALL: haha 04 MOT: han siger mor (.) jeg forstår ikke så meget ham he says mom (.) I don’t understand that much him 05 hvad han siger på tyrkisk til m- til mig what he ((the teacher)) says in Turkish to m- to me 06 MSK: okay 07 MOT: fordi han s- på den måde snakker han (.) because he xxx the way he speaks (.) 08 det er lidt anderledes end vores (.) it is a bit different from ours (.) 09 jeg ved ikke hvor han kommer fra I don’t know where he comes from 10 HDI: han er er he is is 11 MOT: han er ret tydhe is quite xxx 12 HDI: yumuşak g(.) /x:e/ (.) det er /x/ yumuşak g (.) /x:e/ (.) it is /x/ ((he exaggeratingly imitates the teacher’s pronunciation)) 13 XON: det er blødt g it is soft g ((explanation to MSK; then the theme changes)) 89

The sequence illustrates two participants’ evaluations of the teacher’s way of speaking. One is the mother’s judgment in her own words following her contact by telephone (l. 01), the other is Hadi’s report to the mother (l. 04). While both evaluations are associated with a “strange” way of speaking and both were reported by the mother, the mother’s words do not reveal any interruption of understanding between the teacher and the mother. But according to the mother, this is not the case with Hadi, as a pupil in the class, and the teacher’s “strange” variety has impacted the Turkish instruction, at least in the case of Hadi because Hadi does not understand that much of what he says (l. 0405). By characterizing the teacher’s register as “different” from their known and own variety, the mother adds to the value she ascribes the teacher’s significant way of speaking (l. 07-08). At this point, her description becomes identified and specified by Hadi, as if the issue has been depicted beforehand inside the family. Hadi points to what exactly is considered as a criticized feature and exaggeratingly performs the “yumuşak g (.) /x:e/” (l. 12). Hadi’s performance is simultaneous with the mother’s laughter while referring to the teacher’s “strange” way of speaking (l. 01). One can say that this special feature is unexpected, treated as funny, and surprising by the pupils; however, there is no evidence from the fieldwork that the pupils could not understand the teacher. Moreover, the mother cannot identify the teacher’s variety with a region within the Turkish borders: “I don’t know where he comes from” (l. 09). It is, however, difficult to say whether this implies a mis-recognition of the teacher’s register, which might also be localized geographically. Or perhaps the mother in this sequence attempts to stance her and/or her family’s way of speaking and knowing Turkish from the teacher’s variety. Yet what is interesting with this sequence is how linguistic varieties are ideologically focused and represented, and how they give rise to the tension between what register exists, what is preferred (more appropriate or perhaps understandable) inside the MT class, and what is legitimately tied to the geography of Turkey.

90

The teacher accounts for his language variety in a different way. In an interview MSK enquires which type of Turkish he taught, and he answered that the way he pronounced was following the way Kurds acquired Turkish, namely by pronouncing all letters: vi kurdere får den bogstavelige udtalelse (.) for eksempel i Istanbul Istanbul så siger jeg bendigium (.) men vi har lært på den måde (.) som bogstavet siger (.) tyrkisk [we Kurds get the letter-by-letter pronunciation (.) for example in Istanbul Istanbul then I would say bendigium (.) but we have learned Turkish in this way (.) the way the letter sounds ((lit., through a sound-letter correspondence))] (Interview by MSK and XON, 20-022014) In fact, he distinguishes between his variety of Turkish and the variety of Istanbul. In his description of the variety of Istanbul, he says that the Turkish of Istanbul has other pronunciations, and some of the meanings in the lexicon differ from his variety of Turkish. And he adds that based on each variety, you might guess where (which region in Turkey) the speaker comes from. Moreover constructing a symmetry between Turkish varieties (i.e., his vs. the variety of Istanbul) and Danish varieties (the regional vs. Copenhagen variety), he focuses on a linguistic variety hierarchy within the national borders: “It is like Copenhagen dialect that everybody follows it.” Over and above this, Mehmet accounts for his language use and pronunciation rather differently. He adds a pedagogical aspect to the characteristics with which he associates his variety of Turkish. The following example is from the interview MSK and XON conducted with Mehmet, where he describes the register he uses in and around the Turkish class in association with his pedagogy:

91

Example (17): “About the difference of Turkish” Interview with Mehmet (ZGN) by MSK and XON: fieldworkers; 2002-2014. Danish, English. 01 MSK: er der nogle af børnene der har are there any children who find it 02 svært ved at forstå når du difficult to understand when you 03 taler anderledes end de er speak differently from the way they 04 vant til derhjemme tror du are used to ((hear)) at homes in your 05 eller volder det nogen opsigt opinion would it cause any scenes 06 ZGN: jo det tror jeg (.) øh nej yes I think so (.) uhm no 07 det volder ikke så meget (.) it doesn’t cause that much(.) 08 nej det volder ikke så meget (.) no it does not cause that much 09 men nogen gange der er nogle ord but sometimes there are some words 10 som er ligesom øh (.) that are like uhm (.) 11 lytter det samme skriver sound the same as the spelling 12 det samme så nogen gange jeg the same like as sometimes what I 13 trykker på det så jeg siger hvorfor pronounce then I say why 14 du siger det sådan you pronounce it like that 15 hvis du siger xxx den (.) men if you say xxx that (.) but 16 jeg med vilje vil gerne give on purpose I would like to produce 17 den ğ-lyden så du ser ğ-sound then you see 18 hvorfor siger du sådan why you say it that way 19 MSK: forklarer du dem så xxx then do you explain that for them xxx 20 ZGN: jo forklarer jeg så med vilje yes I explain on purpose 21 det er fordi når du skal skrive because when you write down

92

22 23 24 25 26 27 MSK: 28 29 30 ZGN:

du skal ikke lave fejl you shouldn’t make mistake så det er derfor at det der ğ-lyden so that is the reason with ğ-sound jeg trykker på for at det var I pronounce it as it was min mening der skriver my intention when I write jeg læg mærke til ğ-lyden I would pay attention to ğ-sound ja (.) men du forklarer dem ikke yes (.) but you don’t explain them noget om forskellene i tyrkisk about the difference of Turkish sådan inde i Tyrkiet and the like inside Turkey nej no

This excerpt demonstrates that the Turkish teacher is aware of the difference between the Turkish varieties inside the class; that is, the one he teaches in and the one (or the varieties) the pupils use daily. But he insists that his choice of the specific variety of Turkish is conscious (l. 16) and forms part of a pedagogical strategy. In his account, the emphasized pronunciation of the feature soft g helps the pupils to become aware of the fact that soft g (which is unpronounced in other varieties of Turkish) should be there in writing. It also helps them to keep monitoring the difference between the written and spoken language. Thereby they could learn the language accurately and may avoid making literacy mistakes (l. 20–26). According to this account, his insistence on the variant identified with pronounced soft g will result in the fulfillment of the project of literacy. This seems, though, to be Mehmet’s justification of his attitude towards the use of soft g, which is clearly supported by his authoritative role as a teacher. Whereas a challenge exists between the teacher and the pupils, the teacher does not clarify it, keeps using and highlighting it, and ignores protests. What is obvious is that he does not show any interest in relating his applied variety to a certain part of Turkey and comparing that to other existing varieties that may be more recognizable to the pupils (l. 27–30). Although he prior to this

93

sequence had mentioned that one could state precisely which language variety related to what part of Turkey, he did not come up with any metapragmatic explanations for the class; the reason for which remains unclear in the interview sequence. The three examples from the Turkish MT class illustrate that the teacher, being a speaker of a certain variety of Turkish, has chosen a norm of pronunciation that seems to be at odds with the children’s expectations. He also justifies this choice as he presents it as an advantage to the purpose of the class, namely the literacy project. In contrast, the pupils mock his pronunciation of soft g. Also, Hadi in Example (16) reports to his family that the teacher’s way of speaking is “strange” and that “he finds it difficult to understand” his Turkish. Recalling the issue of standardization of Turkish in Turkey, soft g has been one of the salient linguistic features distinguishing the standard and nonstandard varieties. In the case of the study’s data, this feature had particular implications related to literacy in MT classes as well as the recognition of the MT teacher. I will conclude this part with two reflections. First, both teacher and families linked language varieties to certain geographical parts of their country of origin. Their understandings of and ways of connecting language varieties to the geographical regions were independent of what would fit best for the classroom and the pupils’ needs and different language levels. A parent asked where in Turkey the teacher’s Turkish variety might stem from. This question maybe even presupposes the oddness of the Turkish taught in the class. This is because they as Turks are not even able to link it to a concrete region on the Turkish map. Second, the teacher insisted on a variety he identified himself with, used, and at the same time justified. Being aware of different types of speaking Turkish existing in Turkey, his selection indicates his conscious resistance to take the standard and modern variety of Istanbul. While the authority within Turkish and Farsi classes were either the teachers or the parents, we saw a rather different case in the Arabic class. In the next subsection I shall provide evidence from the Arabic group.

94

5.2.3. Standard Arabic In the Arabic classes, the pupils, the teacher, and his assistant teacher were originally from different parts of the Arab world and had access to diverse linguistic resources. This created a linguistic diverse context which is very different from the two former MT classes. Recalling the historical background of Standard Arabic (see 3.2.2.), the varieties related to various Arabic countries, at the level of speaking, form a “continuum of the Arabic dialects,” which are associated with diverse varieties of Ammiya. The Arabic MT class, according to the MT teacher (Aslan in an interview), had as its main goal literacy in Standard Arabic, which differed from Ammiya in terms of lexicon, grammar, and phonetics. This norm was reinforced throughout the materials specified for the purpose of teaching the Arabic MT classes. In a field note a fieldworker notes that the Arabic MT teacher stated that no books existed in dialects such as Lebanese. Thus it is evident that for a mutual understanding between the teachers and the pupils, a variety (or other varieties) of language other than the Fusha was (were) required, or else every participant should know Fusha as a prerequisite to understand the meta-discourse within the instruction. But this was not the case in practice. In addition there was a tendency to use or perhaps to make pupils acquainted with Fusha. My first example in this section deals with this issue. In the sequence in question, Noor who is the assistant teacher in the class helps a pupil, namely Duha (12 y.) with an exercise in her text book. The exercise has to do with vocabulary building and provides more practice with the new lexicon presented in the previous lesson. Duha has to match a set of nouns to a parallel set of adjectives to make noun phrases, i.e., N. + ADJ. The sequence highlights that two varieties of Arabic exist during the interaction: a dialect and Standard; however, when there is a reference to the book the Standard is expected.

95

Example (18): “Al westa” Class recording; Participants: DUH (Duha): pupil & Noor: assistant teacher; 16-05-2013; Arabic, English 01 Noor: ettani Duha next one Duha (.) 02 [shu?] what ((dialect)) 03 DUH: [xxx] alqurun [xxx] ages 04 Noor: alqurun shu? ages what 05 DUH: al wes[ṭa] meddel ((’westa’ with the vowel /e/ is dialect, vs. wusta)) 06 Noor: [alwu]sṭa middle ((’wusta’ with /u/ is Standard Arabic)) Duha is not completely introduced to this exercise prior to this sequence. Noor asks her to take her turn and poses a question in dialect “shu” (l. 02; lit., “what” in Syrian and Lebanese dialect). After having read the first content of the nominal phrase alqurun alwusta (i.e. Middle Ages), she hesitates for a short time to be able to come up with the second part, namely alwusta (l. 03). In response to Noor (l. 04) again in dialect, Duha reads the second part of the nominal phrase, i.e. alwesta (l. 05) in dialect too. Noor in turn corrects her pronunciation (l.06). Now in fact the assistant teacher is at the same time Duha’s mother and shares the same dialect of Lebanese with Duha. But Noor’s attempt to correct the pupil in Standard pronunciation implies there is a preference with reference to the Arabic book. Though, the correction is not accompanied with any metalinguistic accounts. In a similar sequence, the main teacher - Aslan, originally from Iraq - generally used a combination of Standard and Iraqi Arabic. When it came to teaching and in general speaking within the instruction, he showed a tendency to use Standard Arabic. See the following example:

96

Example (19): “Zein mlih” Class recording; Participants: ALN (Aslan): teacher & DUH (Duha): pupil; 7-05-2013. Arabic, Danish, English. 01 DUH: nordvest north west 02 ALN: nord okey north okay 03 zein zein mliḥ good good good ((zein: Iraqi dialect; mliḥ: Standard Arabic)) Example (19) demonstrates that Aslan praises Duha’s correct answer to his prior question, and his compliment is in two varieties, the first one zein is the Iraqi variety, that is, the teacher’s own variety, and the second one mliḥ is the Standard Arabic. This use of both varieties in the same utterance suggests Aslan’s switching from the regional variety to the standard one. Another possible interpretation is that Aslan unconsciously uses the Iraqi variety then self-corrects following the standard one, which is the preferred norm within Arabic instruction institutionally. Even the other examples from the Arabic MT class indicate that the baseline in the communicative situations was Standard Arabic. In addition the language was taught with particular associations. One association dealt with Arabic as lingua franca within the Arab world, and another dealt with the fact that Arabic was not regarded in relation to all Islamic countries. The two ideas were brought up in a session when Aslan initiated an exercise where the pupils should match certain flags to the related Arabic-speaking countries and should further name the related capital cities. During this exercise when a pupil named an Islamic country (which name is unclear in the data), the teacher emphasized that they would not take all the Islamic countries and would select only the Arabic countries instead. Aslan’s point supports that Arabic links the MT classroom to a larger world with Arabic as a shared means of communication. Thereby he maintains focus on an Arabic territory identified with the standard variety that is in line with the discussions on Standard Arabic and the suggested language reforms in Arabic countries, engaging with a more modern Fusha 97

based on the promotion of an Ammiya variety (Haeri 2000). The ideal of an intermediate variety of Arabic understandable for diverse speakers is brought up in what follows. Aslan points to the discrepancy existing between standard and other varieties which pupils bring with them as they are socialized into geographically localized varieties in their families and the potential consequences of the diversity. The following excerpt is part of his account on the consequences of varieties taken from an interview MSK conducted with Aslan. He provides an example indicating what types of challenges might occur as consequences of the varieties’ discrepancies: Example (20): “It is written in the books” Interview with Aslan (ALN): MT teacher by MSK: fieldworker; 24-062013. Arabic, Danish, English. 01 ALN: gardiner som hedder på curtains that you call in 02 standardarabisk som står 03 Standard Arabic as it’s written 04 i bøgerne hedder sidara in the books is called sidara 05 og så siger nej and the ((pupil)) xxx says no 06 det hedder badaja 07 it is badaja 08 og badaja det dialekt and badaja it’s dialect 09 libanesisk dialekt Lebanese dialect 10 fordi mor sir badaja så synes de as the mother says badaja they think 11 badaja det er rigtigt (.) badaja is right (.) 12 æ:h og jeg jeg i jeg m i det kræver u:hm and I I in I mm this requires 13 mere energi fordi jeg har jeg more energy because I have I 14 fokus på bogstavet s som starter I have a focus on the letter s 15 sidara med o:g with which sidara starts a:nd 16 og det skal give mening at: and it should make sense tha:t 98

17 18 19 20 21 22 23 24 25

man bruger sidara i stedet for badaja one uses sidara instead of badaja som mor bruger derhjemme som dialekt as the mother uses at home in dialect det det det her (.) øh men æh it it is it (.) uhm but uhm det kræver også at man finder it also requires that you find et fællesnævner hvor alle a common denominator that everybody ka- nogenlunde at forstå det og ca- to some extent understand it and derfor så bliver jeg nødt til thus then I’m obliged to engang imellem at supplere med dansk once in a while to add Danish når jeg kommunikerer med dem when I communicate with them

In this excerpt Aslan exemplifies how the difference of the varieties, namely the home regional varieties and the educational institution’s variety, makes sense in the class (l. 01–16, 17-18). In his narrative, he depicts a scenario, where a generic pupil with Lebanese linguistic background does not agree with Aslan’s words and thus with the suggested words in Standard Arabic in the book (l. 01-07). The pupil who speaks in Lebanese Arabic (l. 09) has questioned this lexicon. Aslan thinks that this refusing of lexicon is because the pupil has the word from a variety of Arabic she has learnt from her parents as the correct form (l. 10-11). This is daunting to Aslan (l. 12-15), as he has a wordlist to teach and words like “sidara” - with particular constructive letters (that are likely the lesson’s goal) - are named by the pupil as wrong. First, the pupil has changed the focus of teaching by suggesting another word (i.e., “badaja”). In addition, the pupil’s overt rejection of the word to be taught has challenged the correctness and legitimacy of the wordlist in the class material, and thus the teacher’s authority in the class interactional situation. His narrative reveals that Standard Arabic and other nonstandard varities co-exist in relation to participants with Arabic linguistic backgrounds: the home variety and the educational, institutional variety. He also notes that the home-acquired register interferes with his educational aim of uniformity inside the 99

classroom, along with the need to teach and use the non-localized register. In Aslan’s words, the pupil’s rejection entails her parent’s norms and voice “because the mother says badaja, then ((she)) thinks badaja is right.” Although in his narrative, the pupil does not orient to her parents, Aslan interprets her comment as an extension of her parent(s)’ voice (i.e. “this is right”) and dialect (i.e. badaja lexical form). He also considers that this parental voice clashes with the authority of the literacy institution represented by Aslan himself (l. 1-18). The diversity of pupils’ linguistic backgrounds has caused a need for another solution in Aslan’s words - “a common denominator”. In order to obtain pedagogical objectives inside the classroom, the solution has been to use Danish - once in a while (l. 20-25). His idiomatic reference to Danish as a “common denominator” resonates that Danish is shared by all class participants from different linguistic backgrounds, which in turn vary from Standard Arabic. However, it is unclear why Aslan emphasizes the use of Danish as the “common denominator” and not Standard Arabic. Maybe this is to highlight the complicated issue of the variety diversity inside the Arabic MT classrooms that as I discussed above adds to other challenges in teaching them. Or to indicate that MT pupils are not necessarily the presupposed Arabic-speaking children. Or maybe this simply illustrates that Aslan elaborates on the recognition of the Arabic MT classes through displaying them as linked to the ordinary curriculum of the school in Danish. Whatever the reason is, Aslan shows in practice a flexible strategy by using Danish to teach Fusha to the pupils. Whereas Danish functions as a class lingua franca available to all class participants, Fusha implicitly is a lingua franca to be taught. It must be recalled that Fusha is an Arabic lingua franca reinforced by teaching literacy materials produced in line with pro-standard ideology within the Arabspeaking world. I will conclude this chapter with some reflections on the connection between language varieties, language ideologies, and geographical places discussed and observed in and around the three MT classrooms. The adult participants (the teachers, their assistants, and the MT pupils’ parents) come from different geographical regions of their countries of origin. They are also

100

speakers of different varieties of the target languages of the mother tongue classes. For all the adult participants, one may observe a tendency towards one part of the dichotomy of standard variety and diverse regional varieties. For the Turkishand Farsi-speaking adult participants, these varieties are pointed out as concrete regions within their countries of origin (respectively, Turkey and Iran). While the standard variety follows the official standardized register of the countries - with metropolises and institutions as their centers, the regional varieties are tied to other regions with lower socio-political status in the countries. Parents in the study data intended to have their children learn the standard varieties and they associated the geographically localized varieties with devaluing attributes and attitudes. While the norm of teaching inside the Farsi MT class was the standard variety alongside the tendency for a pure language free from foreign words, the selected norm of the Turkish MT teacher deviated from the standard Turkish. He argued for his language use, as a choice and as motivated by the fact that it would give the children easier access to learn Turkish through a pronunciation based on orthography. For the Arabic teacher the defined norm was in line with the norms of the Arab World based on linguistic uniformity. While the standard norm followed the literacy aims of reading and writing in the idealized register of Standard Arabic it did not, though, provide an appropriate interactional space. Standard register varied from the regional varieties pupils were socialized into and brought into the MT classrooms. The standard and the regional in all three cases have historical and sociocultural associations. In other words, the participants’ preference of one register and dis-preference of another may be interpreted as the presence of such associations and historical formations (Agha 2007). The participants’ choice to provide their children with the preferred variety is rooted in a received system of valorization and counter-valorization of the standard and non-standard registers. Yet, the younger participants of the MT classes indicated their sensitivities to different varieties by mocking the teacher in the Turkish case and by addressing their confusion or inability to understand to their parents.

101

6. DISCUSSION AND CONCLUDING REMARKS This study aimed to revisit the concept of “mother tongue” (MT) in the context of Copenhagen, Denmark, and through data collected within and in relation to Mother Tongue Education (MTE). MTE in Denmark is offered by public and private institutions and concerns the teaching of the so-called migrant languages to minority children outside of the ordinary school schedule. This thesis has focused on various understandings of the concepts of MT, MT use, and MTE from different participants’ perspectives in and around MT classes. The study is part of a larger linguistic ethnographic project, and the ethnographic data I have used come from MT classes where Arabic, Farsi, or Turkish is taught. My relation to these classes varied in terms of my role in the field: in the Farsi sites I was an active participant observer; in the Turkish class, I was a passive observer; and as for the Arabic class, my reflections are based on the data other researchers had collected and on the discussions in the research group meetings. My study approached various language ideologies regarding language and space – i.e., interactional/social space and place – to answer: Which space do language users establish through the use of mother tongue? At the outset (2.2.1.), I described language ideologies as a set of normative attitudes towards language and language use. Theories suggest that normative attitudes are formed historically and may involve (1) recognizable patterns of behavior, frequent in relation to a group of individuals; (2) patterns of behavior that are reflexively modeled and assumed as normal among a group of individuals; and (3) patterns of behaviors that are standardized, that is, associated with judgments and valorization of being good, bad, appropriate, and so on (Agha 2007). Then I described the concepts of native speaker, MT user, and heritage learner as a few representations of language ideologies from an academic perspective (2.2.2.). I showed how these concepts were discussed academically, and I demonstrated further 102

how this terminology has been used institutionally and ratified in Denmark with regard to migrant languages. I introduced three types of social agents in and around MT classes: the Danish institutions and authorities, MT parents and teachers, and MT pupils. As for these agents I identified different understandings regarding what is considered as MT, a MT speaker, and the relationship of MT pupils to countries of origin. The MT involved various representations: the way of speaking a given speaker acquires first in relation to the acquisition of other “languages” – e.g. L2; the way of speaking of particular speakers – such as linguistic minority people – as identified with or related to by inheritance links; and the way of speaking associated with particular norms and standards. Based on these understandings, social agents referred to certain pupils as MT pupils in relation to MT classes. The shared aspect of these understandings was the identification of MT pupils’ relation to an ancestral “language” by heritage links. Yet, these links ascertained MT pupils as being related to other languages than Danish in two different ways: (1) they assumed that pupils of a given MT other than Danish were speakers with equal expertise of the alleged MT and met Danish as a foreign language when they entered institutions; or (2) they indirectly linked the pupils to the ancestral languages. The first view was in line with the history of the term MT: a language acquired through a kinship link or the first-acquired language in the context of second language acquisition. The second view was in line with the term heritage language (HL) with a history particularly used in North America. Regardless of the challenges with the term HL (see 2.2.2.), the study’s data illustrated that HL appeared as a more appropriate term in relation to the migrant languages following definitions (Fishman 2001): the HL by definition was not used by pupils as an everyday means of communication; and HL was not taught within the ordinary school curriculum. In addition, the former view linguistically and socially minoritized the pupils. However, the latter view recognized the language and culture of the countries of origin as part of the pupils’ rich repertoire, to which they are probably exposed through

103

their families, and treated the ancestral languages – if acquired by children ever – on an equal footing with Danish. To exemplify the MT participants’ understandings regarding the two linkages mentioned above, I showed that MT teachers were aware of the diversity of MT pupils in terms of proficiency and fluency in the alleged MTs. They also illustrated this diversity during the project fieldwork. According to their accounts, the pupils’ (lack of) proficiency demonstrated their (dis)interest in learning and practicing the MT or possibly a lack of their families’ support for taking MTs as seriously as Danish. The MT teachers also worried about the lack of use of the MT inside the MT classrooms, and they expressed their concern through various strategies, such as overtly saying, “Don’t speak Danish; speak the MT,” claiming a lack of comprehension of Danish and pausing for a short time prior to encourage pupils to shift to the MTs or correcting the pupils. Some patterns of language used by pupils, observed in the three sites, were reported by several parents. In relation to the linkage between language and space, both MT classes and homes were claimed spaces for MTs. In these cases, the concrete MT classrooms and pupils’ homes made a special space within which the claimants (adult participants) demonstrated their linguistic territoriality supported by certain norms of language use and asked the pupils or reminded them of how and what language to (or not to) use. Nonetheless, other cases suggested that, for many parents, it was a regular practice for the children to mix their MT with Danish and to use Danish at home or among their peers and siblings. Both types of teachers and parents reported of their assumptions and preferences concerning MT pupils’ using MTs during family vacations to the countries of origin and took MTs’ use within the countries of origin for granted. MT pupils showed other linguistic behaviors and reacted to the adults’ preferences. In practice, the pupils used the linguistic resources that seemed to fulfill their communicative needs and aims. Also, in fact not all of them could understand the MT or had a large repertoire within it. This of course was at odds with the Danish Ministry of Education’s regulatory definition of the children who were eligible for MTE. In reality, MT pupils in general used Danish as the primary means of

104

communication among siblings and peers as well as in settings where addressees knew Danish, and they also used Danish with the unsatisfied MT teachers and often with their parents. But many of them were aware of their need to communicate with family members outside of Denmark and that this should take place through the MT. With regard to language and space, I posited a further argument that the relationship between MT users and MTs continued further through a linkage between MT pupils and social and geographical places (countries of origin). Therefore, the second part of the study’s theoretical framework, entailing two theories, was presented to situate the study’s second analysis: (1) on the identification of individuals’ selves with respect to others and to a given space and (2) pertaining to partitioning and claiming space, that is, territoriality, being influential in maintaining power, cultural values, and regularity. I used this theoretical approach to explore how the MT was treated by participants in and around an MT class and during interactions and in respect to MT pupils’ countries of origin and to the borders of Denmark. The adult participants partitioned certain places by means of languages to be used (MT, Danish, and English). However, the pupils ignored this spatial partitioning as well as the adults’ claims to these spaces of language use. I argued that MT pupils’ choice of language in interactional situations indexed a protest to authorities – either parental or institutional – and a demonstration of personal space and selves. Yet, this study did not touch on either the impact of the social and institutional hegemony of Danish or the quality of MTE pedagogy on MT pupils’ choice of language; both are worth investigating for more comprehensive insight. I also showed that by virtue of heritage links to the geographical territories – namely, countries of origin, the MT pupils were institutionally positioned both as “bilinguals” and “MT users.” Similar assumptions and positionings were observed in the representations of several adult participants. The two types of assumptions were far from actual practice. The adults also attempted to include the pupils as part of the migrant languages’ speech communities, which were socially constructed. This was

105

not, however, well received by all pupils, and they showed various attitudes towards it. For many of them having a parent from the country of origin was a known fact that did not oppose their relation to Denmark. These pupils suggested their own flexible category, that is, a category containing “both sides” (they identified themselves both as “Dane” and of the country of origin). These cases contrasted two categories of pupils: those who to some extent took the adults’ positioning with respect to the relationship between MT pupil and country of origin and those who rejected being connected to the countries of origin and instead presented a new language–territory norm. For these children living in Denmark could justify the idea of exclusively speaking Danish. The second part of the analyses dealt with linguistic ideologies of standard versus other varieties in relation to MTs. The purpose of this was to investigate what type(s) of language register(s), if any (with a focus on standard or other varieties), pervaded and was preferred in the MT classrooms; how participants identified themselves with these registers; how they represented their dis-alignment with registers; and finally what influence such (dis)preferences had in MT classes. The adults came from different parts of their countries of origin and were speakers of various varieties of the taught languages in MTE classes. For the Farsi- and Turkish-speakers, the dichotomy of standard versus other varieties was situated by the participants within the national borders of the respective countries. This was not the case in Arabic MT classes, as the standard variety is nobody’s MT; rather it is a non-localized linguistic resource inside the Arab world. In the two first sites parents tilted towards the standardized variety of MTs of the countries of origin. While the parents’ preferences of standard variety were in agreement with the taught variety inside the Farsi class, the Turkish MT teacher did not follow the standard variety; thereby, he went against expectations. One Iranian-background parent ideologically associated the standard variety with “normal,” “ordinary,” “formal” and “fluent” language and, regardless of her original variety, dis-aligned herself and her children from geographically localized varieties. A Turkish-origin parent named the variety used and taught in the Turkish MT class as “strange” and difficult to “understand” for the pupils. Her child

106

demonstrated a similar attitude and, like a number of other classmates (observed during the fieldwork), mocked the teacher’s language, taking particular notice of some stigmatized linguistic features. In a step further, the pupil’s parent could not locate the teacher’s variety within the borders of Turkey. On the other side, the Turkish MT teacher insistently used the linguistically significant feature. He claimed it as a pedagogical strategy with benefit for the pupils’ literacy. I provided a different picture in studying the choice between Standard and other regional varieties in the Arabic MT classes. As Standard Arabic always contrasts with the local and regional Arabic vernaculars, the teacher’s tendency to standard variety in MT classes was existentially imposed by the context of Arabic literacy as a language norm originating from a long history of standardization of Arabic in the Arabic countries. This norm was reinforced by the materials chosen for the purpose of teaching in the classes. In a field note a fieldworker notes that the Arabic MT teacher stated that no books existed in dialects such as Lebanese (field note by the researcher Nassri; 20-06-2013). This can be interpreted so that the materials in the Arabic MT classes were selected in line with educational policy in the Arab world. The purpose of this could be the construction of a homogenous speech community of Arabic language users, even outside of the geopolitical territory of the Arab World. Pratt (1987) discussed the notion of speech community in relation to Anderson’s (1983) term of imagined communities, on the one hand, and to linguistics’ subject of study, on the other. She considered the complex notion of speech communities as imagined and defined it by means of three parameters: discreteness, sovereignty, and brotherhood (of members of the community). In Pratt’s perspective, though, such a “view of language” - being carried out in real life - was idealized and “anchored in a normative vision of a unified and homogenous social world” (Pratt 1987: 59). From this perspective and with respect to MT classes, I believe the selection of Standard Arabic is a double-idealization of the “language”; it is idealized diachronically as a lingua franca revealing a means of communication and a particular mode of viewing the world (Pennycook 2012): It is, among others, an index of the Arab world’s identity and

107

brotherhood as the language in its standard variety is imposed to and as the standard disagrees with Arab speakers’ everyday language use. Then standard Arabic is re-idealized inside an MT class in Denmark, where MT pupils cannot necessarily understand their home varieties, do not regularly access Arabic as much as one does in Arabic countries, and are meant to learn Arabic as a means of communication with friends and families. The studies’ two analytical parts addressed the construction of space through the use of mother tongue and by means of ideological understandings. While the first part of the analysis showed how space/place (MT classrooms/countries of origin and Denmark) added to the concrete meaning of alleged MTs, the second part involved how perspectives of the major varieties (standard/non-standard) related to MTs which formed a social space and characterized an institution named the MT class. I suggest that the representation of the relationships between space and language provides a more precise meaning of an institution named “MT class” that is not in agreement with the institutionally established meanings, associations, and references of the concept MT. Particular relationships were taken for granted institutionally concerning MT pupils’ correspondence to MTs and countries of origin and, thereby, defined MT classes as educational places. Yet, the MT pupils in the study constructed other and various relationships to both places and languages, and they did so following interactional means and needs. The diversity of the pupils’ constructed links opposed to and changed the static view of spatial territories, which is in line with what we have inherited from the romantic understanding of a “language” in relation to a “nation” and further to geographical territories.

108

REFERENCES Absillis, Kevin & Jürgen Jaspers (2016): Beware of the weeds: Understanding Flemish linguistic purism as a utopian discourse. International journal of the Sociology of language (forthcoming). Adler, Patricia A. & Peter Adler (1994): Observational techniques. In: N. K. Denzin & Y. S. Lincoln (eds.): Handbook of qualitative research, 377-392. Newbury Park: Sage. Ag, Astrid & Jens Normann Jørgensen (2013): Ideologies, norms, and practices in youth polylanguaging. International Journal of Bilingualism, 17 (4), 525-539. Agar, Michael (1995): Ethnography. In: J. O. Ostman & J. Verschueren (eds.): Handbook of Pragmatics, 583-590. Amsterdam: Benjamins. Agha, Asif (2003): The social life of cultural value. Language & Communication, 23, 231-237. Agha, Asif (2006 [2004]): Registers of Language. In: Alessandro Duranti (ed.): A Companion to Linguistic Anthropology, 23-45. Oxford: Blackwell. Agha, Asif (2007): Language and social relations. Studies in the social and cultural foundations of language. Cambridge, UK & New York: Cambridge University Press. Akinci, Mehmet-Ali (2008): Language Use and Biliteracy Practices of Turkish-speaking Children and Adolescents in France. In: Vally Lytra & J. N. Jørgensen (eds.): Multilingualism and Identities across Contexts. Cross-disciplinary perspectives on Turkish-speaking youth in Europe, 85-108. Copenhagen Studies in Bilingualism, 45. Copenhagen: University of Copenhagen. Albury, N. J. (2016): The power of folk linguistic knowledge in language policy. Language Policy. DOI: 10.1007/s10993-016-9404-4

109

Ammon, Ulrich (2006): Language planning for international scientific communication: An overview of questions and potential solutions. Current Issues in Language Planning, 7 (1), 1-30. Anderson, Benedict R. O’G. (1983): Imagined Communities: Reflections on the Origin and Spread of Nationalism. London: Verso. Appadurai, Arjun (2000): Grassroots globalization and the research imagination. Public Culture, 12 (1), 1-19. Austin, D. E. (2003): Community-based collaborative team ethnography: A community-university-agency partnership. Human Organization 62(2), 143-152. Backus, Ad (2012): Turkish as an Immigrant Language in Europe. In: T. K. Bhata & W. C. Ritchie (eds.): The Handbook of Bilingualism and Multilingualism, 770-779. Oxford: Blackwell Publishing Ltd. Bakhtin, Mikhail M. (1981): Discourse in the novel. In: Michael Holquist (ed.): The dialogic imagination. Four essays by M.M. Bakhtin, 169422. Austin: University of Texas Press. Bale, Jeffrey (2010): International Comparative Perspectives on Heritage Language Education Policy Research. Annual Review of Applied Linguistics, 30, 42-65. Ball, Jessica (2011): Enhancing learning of children from diverse language backgrounds: Mother tongue-based bilingual or multilingual education in early childhood and early primary school years. UNESCO 2010. Blackledge, Adrian & Angela Creese (2010): Multilingualism: A critical perspective. Advances in Sociolinguistics. London: Continuum. Blommaert, Jan (2007): Sociolinguistic scales. Intercultural Pragmatics, 4 (1), 1-19. Blommaert, Jan (2010): The sociolinguistics of globalization. Cambridge: Cambridge University Press. 110

Blommaert, Jan & Jef Verschueren (1998): Debating diversity. Analysing the discourse of tolerance. New York: Routledge. Blommaert, Jan, James Collins & Stef Slembrouck (2005): Spaces of Multilingualism. Language and Communication, 25, 197-216. Blommaert, Jan & Ad Backus (2011): Repertoires revisited: ‘Knowing language’ in superdiversity. Working Papers in Urban Language and Literacies, 67. London: King’s College London. Web: 18th February 2015 Bloomfield, Leonard (1933): Language, revised edition. London: George Allen and Unwin. Böcker, A. (2000): Paving the way to a better future: Turks in the Netherlands. In: H. Vermeulen & R. Penninx (eds.): Immigrant integration: The Dutch case, 153-177. Amsterdam: Het Spinhuis. Chin, Ng Bee & Wigglesworth, Gillian (2007): Bilingualism. An advanced resource book. London & New York: Routledge. Creese, A. (2008): Linguistic ethnography. In: Kendall A. King & Nancy H. Hornberger (eds.): Encyclopedia of language and education, 2nd ed., vol. 10. Research Methods in Language and Education, 229241. New York: Springer. Creese, A., A. Bhatt, N. Bhojani & P. Martin (2008): Fieldnotes in team ethnography: Researching complementary schools. Qualitative Research, 8 (2), 197-215. Cummins, J. (1991): Introduction. The Canadian Modern Language Review, 47 (4), 601-605. Dakwar, R. K. (2005): Children’s attitudes towards the diglossic situation in Arabic and its impact on learning. Languages, Communities, and Education, 2, 75-86.

111

Danmarks Lærerforening (2015): Tosprogede elever – særlig undervisning i dansk som andetsprog. TR-Håndbogen. Web: January 2016 De Swaan, A. (1993): The emergent world language system: An introduction. International Political Science Review, 14 (3), 219-26. Demirci, Mahide & Brian Kleiner (1999): The Perception of Turkish Dialects. In: Dennis R. Preston (ed.): Handbook of Perceptional dialectology, vol. 1, 263-281. Amsterdam / Philadelphia: John Benjamins Publishing Company. Doerfer, Gerhard (1981): Turkische und mongolische Elemente im Neupersischen: Unter besondere Berücksightigung Alterer neupersischer Geschichtsquellen, vor allem der Mangolen- und Timuriden-zeit, 1963-1975. Veröffentlichungen der Orientalischen Kommission, 16, 19-21. Weisbaden: Franz Steiner Verlag. Duchêne, Alexandre, Melissa Moyer & Celia Roberts (2013): Introduction: Recasting Institutions and Work in Multilingual and Transnational Spaces. In: Alexandre Duchêne, Melissa Moyer & Celia Roberts (eds.): Language, Migration and Social Inequalities: A Critical Sociolinguistic Perspective on Institutions and Work, 1-21. Bristol: Multilingual Matters. Duff, P. A. (2008): Heritage language education in Canada. In: D. M. Brinton, O. Kagan, & S. Backus (eds.): Heritage language education: A new field emerging, 71–90. New York & London: Routledge. EACEA P9 Eurydice (2009): Integrating Immigrant Children into Schools in Europe. Brussels: European Commission. Web: January 2016. Elling, Rasmus C. (2013): Minorities in Iran. Nationalism and Ethnicity after Khomeini. New York: Palgrave Macmillan. Erickson, D. & K. Stull (1998): Doing team ethnography. Warning and advice. London: Sage.

112

Fishman, J. A. (2001): 300-plus years of heritage language education in the United States. In: J. K. Peyton, D. A. Ranard, & S. McGinnis (eds.): Heritage languages in America: Preserving a national resource, 81–98. Washington, DC & McHenry, IL: Center for Applied Linguistics & Delta Systems. García, Ofelia (2005): Positioning heritage languages in the United States. The Modern Language Journal, 89, 601-605. García, Ofelia. (2009): Bilingual education in the 21st century: A global perspective. Malden, MA: Wiley Blackwell. Goffman, Erving (1956): The presentation of self in everyday life. Edinburgh: University of Edinburgh. Goffman, Erving (1964): The Neglected Situation. In: P. Giglioli (ed.): Language and Social Contexts, 61-66. London: Penguin. Goffman, Erving (2010 [1971]): Relations in Public. Microstudies of the public order. New Brunswick (NJ.): Transaction Publishers. Gumperz, John J. (1968): The speech Community. In: International Encyclopaedia of the Social Sciences, 381-386. New York: Macmillan. Haeri, Niloufar (2000): Form and ideology: Arabic sociolinguistics and beyond. Annual Review of Anthropology, 61-87. Holmen, Anne & Jens Normann Jørgensen (1990): Modersmålsundervisning. I: Anne Holmen og Jens N. Jørgensen (red.): Modersmålsundervisning i mindretalssprog. Københavnsstudier i tosprogethed, 13. København: Danmarks Lærerhøjskole. Holmen, Anne & Jens Normann Jørgensen (1998): Indledning. I: Anne Holmen og Jens N. Jørgensen (red.): Modersmålsundervisning i mindretalssprog i Danmark. Københavnsstudier i tosprogethed, 29. København: Danmarks Lærerhøjskole.

113

Jernudd Björn H. (1989): The texture of language purism: an introduction. In: B. Jernudd & M.J. Shapiro (eds.): The Politics of Language Purism, 1-20. Berlin: Walter de Gruyter. Jørgensen, Jens N. (2008): Poly-lingual languaging around and among children and adolescents. International Journal of Multilingualism, 5 (3), 161–176. Jørgensen, Jens Normann (2012): Ideologies and norms in language and education policies in Europe and their relationship with everyday language behaviorism. Language, Culture and Curriculum, 25(1), 57-71. Jørgensen, Jens N., Martha S. Karrebæk, Lian Malai Madsen & Janus S. Møller (2011): Polylanguaging in Superdiversity. Diversities, 13(2), 23-37. Web: February 2015. Karrebæk, Martha S. & Narges Ghandchi (2015): ’Pure’ Farsi and political sensitivities: Language and Ideologies in Farsi complementary language classrooms in Denmark. Journal of Sociolinguistics, 19 (1), 62-90. Karrebæk, Martha S. & Narges Ghandchi: The very sensitive question: Chronotopes, insecurities and Farsi heritage language classrooms. Pragmatics and Society. John Benjamins Publishing Company. (Forthcoming). Keyman E. Fuat and Suhnaz Yilmaz (2006): Modernity and Nationalism: Turkey and Iran in Comparative Perspective. In: Gerard Delanty and Krishan Kumar (eds.): The SAGE Handbook of Nations and Nationalism, 425-437. Los Angeles: Sage Publications Ltd. Kristjánsdóttir, Bergthóra S. & Lene Timm (2007): Tvetunget uddannelses politik – dokumentation af etnisk ulighed i folkeskolen. Nyt fra Samfundsvidenskaberne. København: Samfundslitteratur. Kroskrity, Paul V. (ed.) (2000): Regimes of Language. Santa Fe: SAR Press.

114

Kroskrity, Paul V. (2007 [2004]): Language ideologies. In: Alessandro Duranti (ed.): A Companion to Linguistic anthropology, 496-517. Oxford: Blackwell. Langer, N. & W. V. Davies (eds.) (2005): Linguistic purism in the Germanic languages. Berlin / New York: Walter de Gruyter. Lefebvre, H. (2003 [1991]): La production de l’espace. Paris: Anthropus. Leung, Constant, Roxy Harris & Ben Rampton (1997): The Idealized Native Speaker, Reified Ethnicities, and Classroom Realities. TESOL Quarterly, 31 (3), 543-560. Levinson, Stephen C. (1983): Pragmatics. Cambridge: Cambridge University Press. Levinson, Stephen C. (1996): Language and Space. Annual Review of Anthropology, 25, 253-382. Maamouri, M. (1998): Language education and human development: Arabic diglossia and its impact on the quality of education in the Arab region. Discussion paper prepared for The World Bank Mediterranean Development Forum, Marrakesh. Philadelphia: University of Pennsylvania. MacWhinney, Brian (2016): The Talkbank Project. Tools for Analyzing Talk – Electronic Edition. Part 2: The CLAN programs. Web: April 2015 Maegaard, Marie & Pia Quist (2005): Etnografi, praksis og sproglig. Nys – Nydanske Sprogstudier 33, 42-71. København: Dansk Sprognævn. Makoni, Sinfree & Alastair Pennycook (2007): Disinventing and reconstituting languages. In: Sinfree Makoni & Alastair Pennycook (eds.): Disinventing and reconstituting languages, 1-41. Clevedon: Multilingual Matters LTD. Maybin, Janet & Karin Tusting (2011): Linguistic ethnography. In: James Simpson (ed.): The Routledge handbook of applied linguistics. London: Routledge. 115

Milroy, James (2001): Language ideologies and the consequences of standardization. Journal of Sociolinguistics, 5 (4), 530-555. Mæhlum, Brit (2010): Language and social spaces. In: Peter Auer & Jürgen Erich Schmidt (eds.): Language and Space: An International Handbook of Linguistic Variation, vol. 1. Berlin & New York: De Gruyter Mouton. Paikeday, Thomas M. (1985): The Native Speaker is Dead. An informal discussion of a linguistic myth with Noam Chomsky and other linguists, philosophers, psychologists, and lexicographers. Toronto: Paikeday Publishing. Paul, Ludwig (2010): Iranian Language Reform in the Twentieth Century: Did the First Farhangestân (1935-40) Succeed? Journal of Persianate Studies, 3, 78-103. Pedersen, Inge Lise (2009): The social embedding of standard ideology through four hundred years of standardization. In: Marie Maegaard, Frans Gregersen, Pia Quist & Jens N. Jørgensen (eds.): Language attitudes, standardization and language change, 51-58. Oslo: Novus Press. Pennycook, Alastair (2012): Lingua Francas as Language Ideologies. In: Andy Kirkpatrick and Roland Sussex (eds.): English as an International Language in Asia: Implications for Language Education, 137-154. Netherlands: Springer. Perry, John R. (1985): Persian Morphology. In: Aslan S. Kaye (ed.): Morphologies of Asia and Africa, 975-1019. Winona Lake, Ind.: Eisenbrauns. Preston, D. (2005): What is folk linguistics? Why should you care? Lingua Posnaniensis, 47, 143-162.

116

Pratt, Mary Louise (1987): Linguistic Utopias. In: Nigel Fabb, Derek Attridge, Alan Durant & Colin Maccabe (eds.): The linguistics of Writing: Arguments Between Language and Literature, 48-66. New York: Methuen. Rampton, Ben, Janet Maybin, and Celia Roberts (2015): Methodological foundations in linguistic ethnography. In: Julia Snell, Sara Shaw & Fiona Copland (eds.): Linguistic Ethnography: Interdisciplinary Explorations, 14-50. London: Palgrave Macmillan. Retsinformation (2002): Bekendtgørelse om folkeskolens modersmålsundervisning af børn fra medlemsstater i Den Europæiske Union, fra lande, som er omfattet af aftalen om Det Europæiske Økonomiske Samarbejdsområde, samt fra Færøerne og Grønland. Web: January 2016. Sack, Robert D. (1986): Human Territoriality: Its theory and history. Cambridge: Cambridge University Press. Schieffelin, Bambi B., Kathryn A. Woolard & Paul V. Kroskrity (eds.) (1998): Language Ideologies: Practices and Theory. New York & Oxford: Oxford University Press. Silverstein, Michael (1979): Language structure and linguistic ideology. In: R. Clyne, W. Hanks & C. Hofbauer (eds.): The Elements: A parasession on linguistic units and levels, 193-247. Chicago: Chicago Linguistic Society. Silverstein, Michael (1985): Language and the culture of gender: At the intersection of structure, usage, and ideology. In: Elizabeth Mertz & Rick Parmentier (eds.): Semiotic mediation, 219-259. New York: Academic Press. Skutnabb-Kangas, Tove and Robert Phillipson (1989): “Mother Tongue”: The theoretical and sociolinguistical construction of a concept. In Ulrich Ammon (ed.): Status and Function of Languages and Language Varieties, 450-77. Berlin: Walter de Gruyter.

117

Skutnabb-Kangas, Tove (1994): The Debate. Linguistic Human Rights and Minority Education. TESOL QUARTERLY, 28 (3), 625-628. Soja, Edward W. (1971): The Political Organization of Space. Washington D.C.: Association of American Geographer. Spradley, James P. (1980): Participant Observation. Orlando, Florida: Harcourt College Publishers. Statista (2015): Anzahl der türkischen Staatsbürger in Deutchland von 2001 bis 2015. Web: April 2015. Statistikbanken (2016): Web: March 2015 Stern, Hans H. (1983). Fundamental concepts of language teaching. Oxford: Oxford University Press. Storey, David (2001): Territory: the claiming of space. Harlow, UK: Pearson Education Limited. Suleiman, Yasir (2003): The Arabic Language and National Identity. A Study in Ideology. Edinburgh: Edinburgh University Press. The World Bank (2014): Arab World. Web: November 2014 Thomas, G. (1991): Linguistic Purism. London: Longman. Ünal-Logacev, Ö. Fuchs, S. & M. Zygis (2014): Soft ‘g’ in Turkish: evidence for Sound Change in Progress. In: Susanne Fuchs, Martine Grice, Anne Hermes Leonardo Lancia & Doris Mücke (eds.): Proceedings of the 10th International Seminar on Speech Production, 441-444. Working paper 124. Köln: Universität Köln. Undervisningsministeriet (2009): Fælles Mål 2009 – Modersmålsundervisning. Undervisningsministeriets håndbogsserie, nr. 48 – 2009. Undervisningsministeriet.

118

UNESCO (1953): The Use of Vernacular Languages in Education. Monograph on Fundamental Education – VIII. Paris: The United Nations’ Educational and Cultural Organization. UNI·C Statistics & Analysis for the Danish Ministry of Education (2008): Facts and Figures 2007. Key Figures in Education 2007. Statistical Publication no. 3. Web: April 2015 Unipers (2005): The Universal Persian [UniPers] Script. Web: April 2015 United Nations Development Programme (2004): Human Development Report. Cultural Liberty in Todays Diverse World. Web: April 2015 Valdés, Guadalupe (1999): Heritage language students: Profiles and possibilities. IDENTIFIERS, 37-77. Valdés, Guadalupe (2001): Heritage language students: Profiles and possibilities. In: J. K. Peyton, D. A. Ranard, & S. McGinnis (eds.): Heritage languages in America: Presenting a national resource, 3780. Washington DC & McHenry, IL: Center for Applied Linguistics and Delta Systems. Weber, Jean-Jacques & Kristine Horner (2012): Multilingualism. London & New York: Routledge.

Introducing

Wright B. Robin (ed.) (2010): The Iran Primer: Power, Politics, and U.S. Policy. Washington, D.C.: United States Institute of Peace Press. Woolard, Kathryn (1992): Language Ideology: Issues and Approaches. Pragmatics, 2 (3), 235-249.

119

RESUMÉ Dette studie omhandler sprogbrugeres forståelser af sprog og rum i og omkring modersmålsklasser i København. Jeg belyser sproglige ideologier om modersmål, modersmålselev og modersmålsundervisning hos deltagere i sådanne sprogundervisningsklasser. Dermed udforsker jeg, hvordan deltagere forbinder sprog og rum. Dette gør jeg i to analytiske dele. I den første sætter jeg fokus på hvordan deltagernes forståelser for modersmål ideologisk bliver påvirket i forhold til interaktive og sociale rum samt geografiske territorier. I anden del koncentrerer jeg mig om de sproglige ideologier angående standardsprog og dialekter ved at belyse, hvordan deltageres holdninger ideologisk påvirker et socialt sted institutionaliseret som modersmålsundervisningsklasse. Dette synspunkt kaster lys over hvilken sprogtype de voksne deltagere selv anvender eller helst vil have at deres børn tilegner sig og bruger inden for og uden for modersmålsklasserne. Ved at afdække disse to dele besvarer jeg dette spørgsmål: Hvilket rum konstruerer deltagere modersmålsundervisning gennem forståelser?

socialt ud af sprogideologiske

Dette studie indgår i et større sproglig-etnografisk projekt, hvis kerne er modersmålsundervisning i København. Det gør brug af dele af projektets empiri, som blev indsamlet i forbindelse med sprogundervisning i tre mindretalssprog: arabisk, tyrkisk og farsi. Studiets data består af klasseoptagelser, 126 feltnoter, 32 kvalitative interviews med modersmålslærere og nogle forældre til modersmålsundervisningens elever. Jeg anvender sproglig etnografi som metode og analytisk redskab. Jeg anlægger i første omgang et perspektiv ud fra transskriptioner af optagelserne, som jeg underbygger ved dagbøger og refleksioner over disse. Med dette har jeg til formål i anden omgang at opnå en analytikerforståelse af relationer etableret mellem modersmålsbrugere, sociale rum og territorier. Mit studie drager to delkonklusioner. For det første viser studiet, at begreberne 120

vedrørende modersmål, ratificeret af de danske myndigheder i forbindelse med mindretalssprog, tages for givet, men at de absolut ikke passer sammen med den sprogbrug, vi iagttager i modersmålsklasserne. På den anden side identificerer de voksne deltagere modersmålselever som aktivt ikke bruger modersmål og ser dem i relation til forældrenes oprindelseslande. Nogle forældre anerkender en form for fleksibilitet angående elevernes sprogmønstre og positionerer ikke elever som modersmålstalende. Andre definerer begrænsninger i forhold til elevernes sprogbrugsmønstre. Studiet illustrerer yderligere, at eleverne forholder sig forskelligt til disse holdninger, afhængigt af interaktionssituationer. Således viser studiet at modersmålsbegrebet bygger på idealiserede forståelser og anvendes ideologisk. Følgelig kan det give anledning til paradokser i og misfortolkninger af modersmålsundervisning frem for at befordre elevernes sproglige udvikling.

121

APPENDIX (1): TRANSCRIPTION CONVENTIONS [abc]

overlap

xxx

unintelligible

((abc))

comments

(2.0)

pause (time is given in second)

(.)

very short pause

ALL

all participants

UNI

unidentified

˚abc˚

quieter than surrounding talk

LOUD

louder than surrounding talk

↑

emphasis

:

prolongation of sound

haha

laughter

/x/

phonetic reproduction

=

latching

(…)

ellipsis

122

APPENDIX (2): UNI-PERS-ROMANIZATION SYSTEM FOR FARSI

Romanization a â e i o u b p t s j c h x d

Farsi ‫ا‬ ‫آ‬ ‫ ه‬،‫ا‬ ‫ای‬ ‫و‬،‫ا‬ ‫او‬ ‫ب‬ ‫پ‬ ‫ط‬،‫ت‬ ‫ص‬،‫س‬،‫ث‬ ‫ج‬ ‫چ‬ ‫ح‬،‫ه‬ ‫خ‬ ‫د‬

Romanization z r ž š ‘ q f k g l m n v w y

123

Farsi ‫ ظ‬، ‫ ض‬،‫ ز‬،‫ذ‬ ‫ر‬ ‫ژ‬ ‫ش‬ ‫ ع‬،‫ء‬ ‫ق‬،‫غ‬ ‫ف‬ ‫ک‬ ‫گ‬ ‫ل‬ ‫م‬ ‫ن‬ ‫و‬ ‫و‬ ‫ی‬

NOTES

i

I did not have any role in the choice and the organization of the pedagogical content of the classes. ii Yet immigrant languages are called heritage languages in the document of the Euridice (2009) published by the Education, Audiovisual and Cultural Executive Agency – European Commission. Web: http://eacea.ec.europa.eu/Education/eurydice/documents/th ematic_reports/101EN.pdf iii http://www.uvm.dk/Uddannelser/Folkeskolen/Tosprogede/Maalindhold-og-organisering/Modersmaalsundervisning iv https://www.retsinformation.dk/Forms/R0710.aspx?id=23979 v Human Development Report 2004 vi http://tr.dlf.org/Default.asp?AreaID=1006&ChapterID=1072&show=all, under § Modersmålsundervisning: ”Det er foreningens opfattelse, at alle tosprogede elever skal tilbydes modersmålsundervisning, uanset deres oprindelsesland.” vii Ibid. viii While Agha writes ”all other facts of register differentiation”, I prefer to use it in reference to specific geographical bases according to Mæhlum 2010, which is rather in tune with the data in my study. ix Gumperz 1968: 68. x http://www.statistikbanken.dk (the demographic data are from January 2016) xi http://www.denstoredanske.dk/Samfund,_jura_og_politik/Sociolo gi/Grupper/indvandrere xii Maamouri reports that the use of colloquial local variety might indicate the “unacceptable lack of linguistic loyalty” of the language users (e.g., a teacher), which may imply his “treason to ‘Arab Nation’ feeling” (Maamouri 1998: 38). xiii Professor Jens Normann Jørgensen was originally supposed to be part of the Super-MOTE project. He did some initial observations, but then fell terminally ill and had to withdraw. xiv http://childes.psy.cmu.edu/manuals/CLAN.pdf xv http://www.unipers.com/up.htm 124

xvi

Translated from: ”og indbyrdes er deres første sprog det er dansk … det er tydeligt at se (.) om man er tilfreds med det danske sprog”. xvii He says: “de største udfordringer er jo at det er jo faktisk det dominerende store sprog det [er] dansk som man bruger i hverdagen o:g (.) og nogen gange manglende opbakning (.) fra forældreside (.) ø:h og nummer tre er jo også nogle børn som ikke helt (.) efter bogen (.) aktivt talende arabisk derhjemme (.) øh (.) men forældrene kommer fra et arabisk land og det [er] derfor måske nogle gange har de krav på at børnene får undervisning selv om de ikke taler (.) arabisk (.) derhjemme” xviii Translated from: ”de taler arabisk men de taler meget ø:h lokal dialekt eller snæver (.) dialekt som er ikke udbredt og børnene har meget få ordforråd ø:h i hverdagen” xix Translated from: ”de SKAL snakke Farsi hjemme”. xx The data is taken from MSK’s fieldnote (13-02-13): “Der er tydeligvis stor forskel på deres farsi-repertoire. Den lille dreng kan så godt som ingenting, lyder det til, og Narges oversætter hele tiden til dansk. Hans bror er heller ikke så god. MEH har derimod tilsyneladende både langt bedre receptiv og produktiv kompetence.” xxi This was of course not the case with all pupils, and I refer to that in the next set of data. xxii This comment of course differs from the father’s view about Farsi as a MT, and in the case of both Parsa and his brother they were encouraged and sent to the MTE by their Danish mother. xxiii haminjuri is a particular meaning (“of such” or “in this manner”) in this interactional context and adds to the first-personal pronoun “me” in l. 17. Both linguistic features are indexical signs following Peirce’s categorization of linguistic signs (Levinson 1983) and anchors a variety of speaking the language Farsi to the here-and-now context of the interview utterance. But at the same time the mother using the attribute “haminjuri” which originates from an adverb “haminjur” (lit. such) establishes an existential relationship

125

between the way of speaking language known for herself and supposedly (as she might have perceived) for NG. I (NG) have taken “common” as an equivalent for this adjective, which is used mostly in the colloquial language, for the sake of fluency in the translation of the statement.

126